An Analysis of Food Demand Patterns in Hanoi: Predicting the Structural and Qualitative Changes
Mubarik Ali AVRDC – The World Vegetable Center, Shanhua, Taiwan Nguyen The Quan Government Statistics Office, Hanoi, Vietnam Ngo Van Nam Research Institute of Fruits and Vegetables, Hanoi, Vietnam
SUSPER
Sustainable Development of Periurban Agriculture in South-East Asia Project
AVRDC – The World Vegetable Center is an international not-forprofit organization committed to alleviating poverty and malnutrition through research, development, and training.
AVRDC – The World Vegetable Center P.O. Box 42, Shanhua, Tainan, Taiwan 74199, ROC tel: +886-6-583-7801 fax: +886-6-583-0009 e-mail: avrdcbox@avrdc.org web: www.avrdc.org © 2006 AVRDC – The World Vegetable Center ISBN 92-9058-145-X Edited by Tom Kalb; cover design and photo lay-out by Ming-che Chen Citation Ali, M., T.Q. Nguyen, and V.N. Ngo. 2006. An analysis of food demand patterns in Hanoi: predicting the structural and qualitative changes. Technical Bulletin No. 35. AVRDC publication number 06-671. Shanhua, Taiwan: AVRDC – The World Vegetable Center. 62 pp.
Contents
i
Contents
Chapters Tables Figure Foreword Acknowledgements i ii iii iv v
Chapters
1 2 Introduction Survey Methodology 2.1 Sample selection 2.2 Data collection Analytical Procedure 3.1 Indicator definitions 3.2 Comparative test 3.3 Estimation of food diversity 3.4 Estimation of elasticities Results and Discussion 4.1 Characteristics of the sample 4.2 Average consumption pattern 4.3 Consumption by location 4.4 Consumption by income group 4.5 Consumption by farm type 4.6 Consumption by season 4.7 Sources of food 4.8 Farming based group 4.9 Food quality 4.10 Demand elasticities Summary and Policy Implications 1 3 3 4 5 5 6 6 7 8 8 9 11 14 17 20 24 27 28 38 42 45 49
3
4
5
References Appendixes
ii
Contents
Tables
Table 1. Table 2. Table 3. Table 4. Table 5. Table 6. Table 7. Table 8. Table 9. Table 10. Table 11. Table 12. Table 13. Table 14. Table 15. Table 16. Table 17. Table 18. Table 19. Table 20. Table 21. Table 22. Table 23. Table 24. Table 25. Table 26. Table 27. Table 28. Table 29. Table 30. Table 31. Distribution of sample by income group and region General characteristics of surveyed household by location Education and occupation of household member by location Per capita daily food consumption and expenditure Per capita daily consumption of vegetables and fruits and expenditure share Per capita daily consumption of food (g) by food group and location Per capita daily expenditure of food (VND) by food group and location Average price of food (1000 VND/kg) by food group and location Per capita daily consumption of vegetables and fruits by their group and location Average price of vegetables and fruits (1000 VND/kg) by group and location Per capita daily consumption of food (g) by food and income groups Per capita daily expenditure of food (VND) by food group and income group Average price of food (1000 VND/kg) by food group and income group Per capita daily consumption of vegetables and fruits (g) by their group and income group Per capita daily expenditure on vegetables and fruits (VND) by income groups Average prices of fruits and vegetables (1000VND/kg) by income groups Per capita daily consumption of food (g) by food group and farm type Per capita daily expenditure of food (VND) by food group and farm type Per capita daily consumption and percent share of vegetables and fruits by their group and farm type Per capita daily expenditure (VND) on fruit and vegetable group by farm type Average price of vegetable and fruits group (1000VND/kg) by farm type Per capital daily consumption of food (g) by food group and season Average price of food (1000VND/kg) by food group and season Per capital daily expenditure of food (VND) by food group and season Per capital daily consumption (g) of vegetable and fruits group by season Average price of vegetables and fruits (1000VND/kg) by season Per capital daily expenditure (VND) on vegetable and fruit groups by season Source of food (% of quantity) by group Food sources (% of quantity) by region and food group Food sources (% of quantity) by income and food groups Food sources (% of quantity) by farming based and food groups 5 8 9 9 10 11 12 12 13 13 14 15 15 16 16 17 18 19 19 20 20 21 22 22 23 23 23 24 25 27 28
Contents
iii
Table 32. Daily per capita availability and deficiency level of major and micronutrients Table 33. Nutrient source by food group Table 34. Daily per capita availability and deficiency level of major and micronutrients Table 35. Daily per capita availability and deficiency level of major and micronutrients by location Table 36. Relationship of food diversity and food and nutrient consumption Table 37. Food diversity by location, and respondent group Table 38. Processing stage of food (% of total food) by food group and region Table 39. Processing stage of food (% of total food) by food group and region Table 40. Processing stage of food (% of total) by food group and income level Table 41. Processing stage of food (% of total food) consumed by food group and professional group Table 42. Food eaten outside (%) by region and income and farmer group Table 43. Results of the analysis of the price difference across region and income group (number of commodities) Table 44. Regression analysis of food quality as represented by price difference (%) across region and income group Table 45. Demand elasticities of major food groups in Hanoi Table 46. Own-price demand elasticities of major food groups by location and income group in Hanoi Table 47. Demand elasticities by vegetable group
29 30 31 32 32 33 34 34 35 35 36 37 37 38 40 41
Figure
Fig. 1. Seasonal share in annual expenditure 21
iv
Foreword
Foreword
Currently about one-half of the world’s population of 6 billion reside in cities, and by 2030 the world’s population is predicted to grow to 8 billion with 60% living in urban areas. With this increase in population as well as enhanced income, the food requirements of urban residents will be enormously increased both in terms of quantity and quality. To meet these changing food needs of cities requires enormous planning, which depends upon an accurate knowledge of the consumption patterns of the dwellers and the sources of their food supply. This report provides such a comprehensive analysis for the city of Hanoi, Vietnam. The report not only analyzes the changes in dietary patterns that are expected to occur with urbanization and enhanced income, it pays special attention to the differences in food in terms of nutritional composition, diversity, prices, processing stage, and extent varieties are eaten across various population groups. Although changes in food consumption habits of city people are constantly evolving, a quantitative analysis of the changes is lacking. The analysis in this report fills this gap. The food requirements in urban areas are not only complex, but these requirements are dependent upon the policy environments in city areas. Different socioeconomic groups in the city may have different food situations to start with, and they may respond differently to the changes in city policy environments. This report captures these differences by conducting the analysis for different socioeconomic and regional groups, and by estimating food demand elasticities for these groups in Hanoi. The estimates of demand elasticities provide an innovative tool for predicting the changes in consumption patterns without repeating the household consumption surveys, which is an expensive undertaking. The analysis was conducted from household survey data collected by the authors themselves. Unlike the once-a-year household surveys conducted by the national statistical bureaus, this survey was conducted three times in a year. This was a massive undertaking, but enabled the authors to capture seasonality in the consumption pattern, especially in fruits and vegetables. Moreover, the role of these fruits and vegetables in supplying nutrients is very well defined in this analysis. I sincerely hope that this analysis of Hanoi will serve as a model in predicting the quantitative and qualitative changes in the consumption patterns of other cities, which then will become the basis of innovative policy planning for meeting the food requirements of growing cities.
Thomas A. Lumpkin Director General of AVRDC – The World Vegetable Center
Acknowledgements
v
Acknowledgements
This study was conducted under Component II of the Sustainable Development of Periurban Agriculture in South-East Asia Project (SUSPER), which is financed by the French Ministry of Foreign Affairs (FMOFA). We would like to express sincere thanks to the FMOFA for its financial support for this research. In the project office, Dr. Hubert de Bon, Dr. Paule Moustier, Mr. Boun-Tieng Ly, Dr. Isabelle Vagneron and Ms. Nguyen Thi Thu Nhai lent their strong administrative support to the research which we acknowledge with thanks. A special thanks to Dr. Hubert de Bon for his consultation and coordination in developing a design framework for research. Acknowledgements are due to Mr. Hoang Bang An, Mr. Trinh Quang Thoai, Ms. Le Thuy Hang, Mr. Dang Dinh Dam, and Mr. Le Nhu Thinh of the Research Institute of Fruits and Vegetables (RIFAV), as well as Mr. Nguyen Phong and Mr. Nguyen Van Dong of the General Statistics Office (GSO) for their help in the field work. Thanks are extended to staffs of the three provincial statistics offices and the district statistics departments in Hanoi, Hung Yen and Ha Tay provinces. The authors especially thank the hundreds of household heads who shared information about their diet, income and family structure with the survey team. The authors also acknowledge the help of Dr. Umar Farooq in providing guidance during the field work, and Ms. Mei-huey Wu for assisting in the analysis. The help of Ms. Olivia Liang for typing and layout is also appreciated. We would like to thank Dr. Thomas Kalb for editing as well as Mr. Ming-che Chen for cover design.
Introduction
1 Introduction
The food requirements of Hanoi have enormously increased in recent years both in terms of quantity and quality, and will continue doing so in the near future. To meet the changing food needs of the city requires enormous planning, which depends upon an accurate knowledge of the consumption patterns of the dwellers and the sources of their food supply. The gap in consumption patterns across various socioeconomic and regional groups in and around the city and seasonal differences in food demand have increased the need of such analysis. This study provides a detailed quantitative description of the consumption patterns of Hanoi residents across various income and regional groups and compares it with its rural surroundings. We also describe various food sources within and around the city. The food quality across various population groups is analyzed in terms of its difference in nutritional composition, diversity, prices, processing stage, and extent of it eaten outside the home. Despite the importance of tracking down the changing food demand pattern for effective planning, it is costly to estimate the consumption patterns regularly based on household consumption surveys. To overcome this problem, the demand and income elasticities of different food items are used to predict the demand for these items in the near future. This study provides a detailed estimate of food demand and income elasticities for various food groups. These elasticities, in addition to predicting the future food demand, can also be used in estimating the welfare impact of various technological changes in food production and marketing. Moreover, these elasticities can be applied in simulating the impact of various price policies and market shocks in the domestic and international markets. The new definition of food security adopted by the International Conference on Nutrition in 1992 emphasizes the access to nutritious food, rather than simply meeting the protein energy-requirements (FAO, 1996). This implies that, for a healthy diet, consumers should have physical and economic access to the food that not only provides a minimum protein-energy requirement but contains certain levels of necessary micronutrients as well. This study analyzes this aspect of food quality across various population groups and regions. The knowledge of various nutrient sources and nutrient deficiency in the existing consumption patterns can help policy planners in designing policies to mitigate these deficiencies and meet the future nutrient demands of the population. A growing body of literature is emerging signifying the effective role of vegetables in eradicating micronutrient deficiency problems.1 Moreover, vegetables are the main source of food diversity (Ali and Farooq, 2004). Therefore, in this study we emphasized on the consumption pattern of vegetables in more detail. The demand and income elasticities are also estimated for main vegetable groups, which to our knowledge is the first attempt of its kind for Hanoi.
1
For the details of these studies, see Farooq and Ali (2003).
An Analysis of Food Demand Patterns in Hanoi
The data for this study were collected through a household consumption survey using the 24-hr recall method. The survey was conducted in the urban and peri-urban areas of Hanoi, and in rural provinces around Hanoi. To cover seasonality in fruit and vegetable consumption, the survey was repeated three times in a year, representing three distinct seasons. Appropriate representation was given to different income groups in the survey. In this way, this study gave us a unique opportunity to compare the consumption pattern across various regions, income groups, and seasons, as well as provided us with a broad-based data to estimate demand and income elasticities for different food groups, especially vegetables.
Survey Methodology
2 Survey Methodology
2.1 Sample selection
One of the objectives of the study was to compare the food consumption patterns in urban, peri-urban, and rural areas near Hanoi. To represent the urban and peri-urban Hanoi, we took all of its seven urban and five rural districts, respectively. To represent rural areas, we selected Hatay and Hungyen provinces. The household sample selection in this study came from the sample frame of the Vietnam Household Living Standards Survey (VHLSS2002) conducted by General Statistics Office of Vietnam (GSO) during 2002. The VHLSS2002 surveyed 60,000 households living in 3,000 enumeration areas (EA) in 3,000 communes/wards throughout the country during the first six months of 2002.2 Following three steps were used to select households in and around Hanoi from the VHLSS2002: Step 1: Selection of districts. We selected all districts of urban and peri-urban Hanoi, and randomly selected three districts from fourteen rural districts of Hatay and three districts from nine rural districts of Hungyen (see Appendix 1 for the provinces and districts included in the sample). Step 2: Selection of EAs (or communes/wards). We randomly selected: Urban area of Hanoi: Rural area of Hanoi: Rural area of Hatay: Rural area of Hungyen: 25 EAs from 48 EAs 25 EAs from 32 EAs 15 EAs from 17 EAs 15 EAs from 18 EAs
(See Appendix 1 for the communes/wards included in the sample) Step 3: Selection of households. We randomly selected 10 households from the 20 households interviewed in the VHLSS2002 in each selected EA. If selected households did not co-operate or no longer lived in the selected EAs, they were replaced with households from the 10 remaining households in each EA. In this way, we selected a total of 800 households for our survey. The regional distribution of this sample was as follows: 250 households of urban area of Hanoi province, 250 households of rural area of Hanoi province (peri-urban area), 150 households of rural area of Hatay and 150 households of rural area of Hungyen provinces. The selected household remained the same throughout the three rounds. In this way, we conducted a total of 2,400 interviews in the three rounds.
2
The VHLSS2002 was a one-point survey, and therefore, did not cover the seasonality of food sources. Moreover, this analysis was conducted nationwide and not comprehensively for Hanoi.
An Analysis of Food Demand Patterns in Hanoi
2.2 Data collection
The data was collected through interviews with housewives, household heads, or other people in the family engaged in cooking. For this purpose a questionnaire was developed by AVRDC – The World Vegetable Center, and pre-tested in Hanoi peri-urban in the hot-wet season of 2002. It was modified to meet the objectives more precisely. The survey team consisted of researchers from the Research Institute of Fruits and Vegetables (RIFAV) and GSO. Field enumerators were taken from RIFAV and field supervisors from GSO. They were trained by a socioeconomist from AVRDC. Six types of information were obtained through each interview: 1) characteristics of household, such as age, occupation, education of every household member, etc.; 2) size of farm and home garden and outputs harvested at the time of survey; 3) food quantity of individual food items consumed in the family in the latest three meals at home and family members who participated in each meal; 4) source of each food item; 5) processing stage of the food; and 6) price or total cost of each food item.
Analytical Procedure
3 Analytical Procedure
3.1 Indicator definitions
Results in this report are discussed by location, income group, farm type, season, food group, vegetable and fruit groups, and nutrients. These are explained as follows: Location. Three locations were defined: urban, peri-urban, and rural. “Urban” includes households selected from the urban districts of Hanoi, “peri-urban” includes households from rural districts of Hanoi, and “rural” includes households from rural areas of Hatay and Hungyen (Appendix 1). Income. Groups were defined as low, low-mid, middle, mid-upper, and upper on the basis of the income classification used in the VHLSS2002. The ranges of monthly per capita income for each income group were separately defined for each surveyed provinces (Table 1). Table 1. Distribution of sample by income group and region
Income group Low Low-mid Middle Mid-upper Upper Total Hanoi Per capita income range Sample size (000 Urban Peri-urban VND/month) (no.) (no.) < 296.9 296.9–430.1 430.2–569.9 570–818.7 > 818.7 16 43 66 61 64 250 106 56 47 27 14 250 Hatay Per capita income range Sample (000 size VND/month) (no.) < 136.6 136.6–190.4 190.5–255.3 255.4–373.6 > 373.6 33 31 26 32 28 150 Hungyen Per capita income range Sample (000 size VND/month) (no.) < 155.2 155.2–202.4 202.5–261.9 262.0–370.7 > 370.7 29 21 29 41 30 150
Farm type. Farms were classified as vegetable farmer, non-vegetable farmer, nonfarmer in urban, non-farmer in peri-urban, and non-farmer in rural groups. Out of the total of 2,400 households surveyed in the three rounds, 1,328 were farmers (407 vegetable and 921 non-vegetable farmers), and 1,072 were non-farmers (54 in rural areas, 293 in peri-urban areas, and 725 non-farmers were in urban areas). Season. This survey was carried out in three rounds: October–November (cold-wet season), February–March (cold-dry season) and June–July (hot-wet season). Food and vegetable groups. Seven food groups were defined: cereals, vegetables, fruits, meats, aquatic products, egg and milk, and others. Seven vegetable groups were defined: allium, other root and stem, heading cole, cucurbits, other fruit and flower, leafy, and pulses. For the construction of these groups, see Appendix 2.
An Analysis of Food Demand Patterns in Hanoi
Nutrient group. The data of each food item consumed in each household were converted into nutrients using the Food Composition Table from Vietnam published by National Nutritional Institute in Hanoi during 1999. If any nutrients of a food were not available from this source, it was taken from FIRDI and PUST (1998). Nine nutritional components were considered important in this study: calories, protein, calcium, iron, vitamin A, vitamin B1, vitamin B2, niacin, and vitamin C. The calories were separately estimated for fat and nonfat sources using the conversion factors reported for various food groups in Merrill and Watt (1995). Food prices. Food prices were estimated as the cost on each item divided by its quantity consumed. In case of home-produced food, we assumed the average shadow prices for all households in the same commune who bought that food item. Food quality. Food quality was measured in terms of difference in its in nutritional composition, diversity, prices, processing stage, and extent of it eaten outside. Food consumed outside home. It was difficult to estimate the quantity and expenditure of food eaten outside using the 24-hr recall method. This is because of the absence of the family members from the house who have eaten food outside at the time of the survey. We overcome this problem by assuming that the persons missing from a meal in the family eat outside, and they eat the same food and similar food quantities as they eat at the house. The estimated food quantity eaten outside was then added in the quantity prepared inside the family to estimate the percentage of food eaten outside. Although there is no reason to believe that the structure of the food eaten outside in general will be different than the structure of the food prepared at home, the estimates will be true to the extent that these two structures deviate from each other. This percentage of food eaten outside will be underestimated due to the sample selection bias against those families whose all the members have eaten all the three meals outside, thus have no probability of being included in the survey, although such probability in our sample was small.
3.2 Comparative test
The parameter values across groups were tested using the paired t-test with unequal variance .
3.3 Estimation of food diversity
The most commonly used diversity indices are some special cases of the following form (Hannah and Kay, 1977):
DTF = ∑ ( S ia )1 /(1−a )
where DTF is the food diversity index, Si is the share of the ith item in total food, and α is the diversity parameter, such that α > 0 and α 1. For α = 2, the index becomes the ≠ inverse of the Herfindahl-Index that is commonly used to measure industry concentration (Escalante and Barry, 2001; Hanson and Simons, 1995).
i −1
m
Analytical Procedure
The general index measures both the number of items and the evenness of item shares, with the parameter a determining the weight of number of items versus evenness. The higher the α value, the greater the emphasis on the evenness, while a parameter value of α = 0 simply counts the number of items (Tauer and Seleka, 1994). The upper limit value of the index for any α value is the number of items, and the lowest limit is 1. In this study, the inverse of Herfindahl-Index was used, and was measured from the quantities of individual food items.
3.4 Estimation of elasticities
Alternative demand elasticity estimation procedures and their rationales are discussed in Appendix 3. The demand elasticities were estimated using the complete demand system in which all products are treated symmetrically and simultaneously. To avoid the large number of parameters involved in estimating a complete demand system, we used a twostage budgeting analysis. This procedure assumes that consumer’s budget allocation for food can be decomposed into two separate stages. In the first stage, total expenditure is allocated over broad food groups. At the second stage, expenditure is further allocated over its sub-groups. The Linear Expenditure System (LES) or Almost Ideal Demand System (AID) model is used at the first stage, while only the AID model is applied at the second stage (Appendix 3, Table A1). We tried using both AID and LES models at the first stage and selected the best model based on consistency of results. Estimating demand elasticities for individual food items/sub-groups using the crosssection household survey data, a common problem incurred because of zero consumption of various food items by many households. Heien and Wessells (1990) proposed a twostep estimation procedure to resolve this problem. However, Shonkwiler and Yen (1999) pointed out an internal inconsistency in the estimators obtained using this procedure. They proposed a slightly different two-step estimation procedure, which was used in this study (see Appendix 3 for details). The first stage in the two-stage budgeting procedure was applied to seven food groups, while the second stage was also applied to seven vegetable subgroups defined earlier. The linear expenditure system equation was estimated in shares form, because this specification is less likely to involve heteroskedasticity than expenditures (Pollak and Wales, 1978). Use of the household characteristics in the share equation also helped to mitigate this effect.
An Analysis of Food Demand Patterns in Hanoi
4 Results and Discussion
4.1 Characteristics of the sample
The family size does not vary across the three sites. The areas under total house and under covered-house get larger as we move from urban to rural areas. The size of farm in the peri-urban area is many times larger compared to the farm size in the urban area, but less than half the size of farm found in rural areas. Similarly garden size gets bigger as we move from urban to rural areas. Farm area is mostly owned by the family and very little is rented or borrowed from other families. Private and public tap is the major source of drinking water in urban Hanoi, while peri-urban Hanoi mainly depends upon underground water pumped by the family owned tubewell. In rural areas, main sources of drinking water are underground water (pumped by family pump), rainwater, and wells. Contamination of rainwater may be a problem, as it is stored in open pounds. Threefourths of households in urban areas and over one-fourth of households in peri-urban areas owned refrigerators, while only 4% of the rural households owned a refrigerator (Table 2). Table 2. General characteristics of surveyed household by location
Character Urban Peri-urban 4.2 a 279.2 b 84.5 b 904.7 b 111.1 b 98.0 0.8 1.2 8.4 10.8 0.0 64.0 14.8 2.0 0.0 27.2 Rural 4.3 a 387.3 a 91.8 a 2,009.2 a 186.9 a 97.0 1.3 1.7 1.0 0.7 0.3 43.3 19.0 35.3 0.3 4.0 Average 4.2 249.1 74.4 1,045.3 105.8 95.8 1.1 3.1 19.2 15.9 0.3 38.9 11.8 13.9 0.1 33.2 Number of person in a family 4.0 a House area (m2) 52.5 c Cover area (m2) 43.2 c Farm area (m2) 25.2 c Garden area (m2) 2.6 c Ownership of household (%) Owned 92.0 Rented 1.2 Borrowed 6.8 Main source of drinking water (%) Private tap 51.8 Public tap 39.4 Purchased in container 0.4 Underground water pumped by family pump 8.4 Constructed well 0.0 Rainwater 0.0 River/lake/pond water 0.0 Household having refrigerator (%) 74.3
Mean separation in rows by Duncan’s multiple range test at P < 0.10.
On average, about one-fourth of the sample household members are illiterate or achieved only the primary level of education. The proportion of persons who are illiterate, or achieved only the primary or lower secondary levels of education increases as we
Results and Discussion
move from urban to rural areas, but the trend is opposite for upper secondary, technical worker, college, university, or PhD degree holders (Table 3). Table 3. Education and occupation of household member by location
Degree/occupation Degree/certificate/diploma (% of farmers) Illiterate or below primary Primary Lower secondary Upper secondary Technical worker/Intermediate/ Professional secondary College/University Master/Ph.D./Professor Schooling year (number) Occupation group (% of farmers) Unemployed Leaders in all fields and levels Professionals in all fields Personal services, protection and sales Skilled manual workers Assemblers and machine operators Unskilled workers Others Urban 17.7 11.7 23.2 23.4 7.2 15.7 1.1 10.1 11.0 0.0 42.8 24.0 10.8 3.1 7.5 0.8 Peri-urban 23.5 20.1 29.8 18.1 4.4 4.1 0.1 8.1 3.5 0.5 15.5 11.3 14.5 2.0 51.4 1.3 Rural 28.0 23.9 32.9 11.4 2.2 1.7 0.0 7.1 3.0 0.4 5.1 7.2 7.0 0.8 75.6 0.9 Total 23.5 19.1 29.0 17.0 4.4 6.6 0.4 8.3 5.2 0.3 18.1 12.8 10.4 1.8 50.4 1.0
4.2 Average consumption pattern
In Hanoi and surrounding rural areas, an average of 1 kg of food per person is consumed daily and about 6,500 VND per person is spent on food everyday. Cereals make up about 39% of the total quantity of food consumed in the diet (Table 4). Vegetables are the second highest component of diet in terms of quantity contributing about 24% Table 4. Per capita daily food consumption and expenditure
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Quantity (g) 411.7 253.4 90.4 92.2 46.2 22.1 129.6 1,045.7 Food Share (%) 39.4 24.2 8.6 8.8 4.4 2.1 12.4 100.0 Food expenditure Expenditure (VND) Share (%) 1,651.0 532.3 502.8 2,099.8 587.2 367.8 731.7 6,472.5 25.5 8.2 7.8 32.4 9.1 5.7 11.3 100.0
0
An Analysis of Food Demand Patterns in Hanoi
share. Fruits and meats each claim about 9% share in the diet. A large share of food (about 12%) goes to small food items (Table 4). These results are consistent with those reported in Thuy et al. (2002). In terms of food expenditure, however, meats claim the largest share followed by cereals together claiming more than half of the food expenditure. Vegetables, fruits, and aquatic products each claim 8–9% share, and egg and milk about 6% share in total expenditure. Other miscellaneous food items claim 11% of the food budget share (Table 4). About 253 g of vegetables per person is consumed daily and 532 VND is spent on vegetables in Hanoi and its surrounding rural areas. Leafy is the major vegetable type both in terms of total quantity and expenditure. Heading cole and other stem and root are the next important vegetables. Expenditures on vegetables and fruits are nearly equal even though the average daily quantity of vegetables consumed is more than double that of fruits (Table 5).
Table 5. Per capita daily consumption of vegetables and fruits and expenditure share
Vegetable/fruit Vegetable Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Total vegetable Fruit Tropical fruits Sub-tropical fruits Temperate fruits Total fruit Quantity Quantity (g) Share (%) 7.4 28.6 37.5 21.3 14.6 131.0 12.9 253.4 56.2 26.8 7.4 90.4 2.9 11.3 14.8 8.4 5.8 51.7 5.1 100.0 62.1 29.7 8.2 100.0 Expenditure Expenditure (VND) Share (%) 24.6 66.7 61.0 43.3 45.1 235.2 56.3 532.3 219.2 229.5 54.1 502.8 4.6 12.5 11.5 8.1 8.5 44.2 10.6 100.0 43.6 45.6 10.8 100.0
The tropical type are the major fruits consumed in Hanoi and its surrounding rural areas, claiming about 62% of total quantity and 44% of total expenditure on all fruits. The tropical location of Vietnam and its favorable conditions for diversified tropical fruit production contribute to these results. The sub-tropical fruits are more expensive and less available and consumed. Therefore, even though the total quantity of sub-tropical fruits consumed is less than tropical fruits, expenditure and budget share on both are similar. The quantity of temperate fruits consumed and expenditure made on these are relatively small (Table 5).
Results and Discussion
The average consumption pattern in Hanoi and its surrounding areas hides variations across various socioeconomic groups of the population. The following sub-sections describe these variations in consumption patterns.
4.3 Consumption by location
The total quantity of per capita daily food consumed increases as we move from rural to urban areas (Table 6). This can be explained in terms of income trend in the same direction. Moreover, flow of food from various production areas make many kinds of food available throughout the year in urban areas while relatively poor links with other food production centers limit the availability of food in the rural area. Not only is the total quantity of food consumed higher, the quality of food measured in terms of the consumption of vegetables, fruits, meats, egg and milk is also better in urban compared to rural areas (Table 6). Part of this increase came from the reduction in the consumption of cereals, as urban people regarded cereals as low-quality food, and partly it is because of higher income. Consumption of aquatic products is statistically not different across the two regions but surprisingly low in the peri-urban areas. Table 6. Per capita daily consumption of food (g) by food group and location
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Rural 442.2 a 236.7 c 37.9 c 76.7 c 50.3 a 14.9 c 126.2 b 985.0 c Peri-urban 424.8 b 258.1 b 69.5 b 86.8 b 36.9 b 20.0 b 140.0 a 1,036.0 b Urban 362.1 c 268.8 a 174.5 a 116.4 a 50.5 a 32.9 a 123.4 b 1,128.5 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
The difference in expenditure on all kinds of food between urban area compared to peri-urban and rural areas is even more significant than total quantity of food (Table 7). This can be explained in terms of higher prices of food in the urban area (Table 8). Infrastructure, especially the market system, of the peri-urban area is better than that of the rural area but not as good as those of urban area; therefore quantity, quality and prices of food items consumed by people living in peri-urban areas usually are intermediate between those in urban and rural areas. The average prices of different food groups generally increases significantly as we move from rural to peri-urban and urban areas, although egg and milk is the exception. The difference is generally more startling between rural and urban areas ranging from 14% for egg and milk and 127% for aquatic products (Table 8). Some of the difference (2–5%) may be attributed to the difference in transportation cost from farm to these markets and some to structural differences of food groups in each region. We believe,
An Analysis of Food Demand Patterns in Hanoi
however, that major differences are attributed to value addition, such as grading and cleaning, and to better quality of the food item itself.
Table 7. Per capita daily expenditure of food (VND) by food group and location
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Rural 1,394.8 c 373.5 c 122.8 c 1,425.3 c 438.9 b 240.0 c 613.3 c 4,608.6 c Peri-urban 1,659.0 b 506.7 b 348.9 b 1,892.2 b 452.7 b 306.0 b 718.6 b 5,884.1 b Urban 1,952.7 a 750.0 a 1,117.3 a 3,124.6 a 902.1 a 584.8 a 888.0 a 9,319.4 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Table 8. Average price of food (1000 VND/kg) by food group and location
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Rural 3.2 c 1.6 c 3.5 c 16.1 c 10.0 c 19.8 a 6.8 c Peri-urban 4.1 b 2.0 b 5.1 b 19.7 b 15.2 b 16.8 b 7.2 b Urban 5.6 a 2.9 a 6.3 a 26.6 a 22.7 a 22.6 a 9.6 a Overall 4.2 2.2 5.2 20.5 16.1 20.0 7.8
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Looking at individual vegetable consumption pattern, leafy types contribute about one-half of the total quantities of vegetable groups consumed in each of the three sample areas (Table 9). The consumption trends of heading cole, other fruits and flowers, and pulses across regions are the same as those of total vegetables (i.e., consumption increases as we move from rural to urban areas although heading cole is not significantly different across urban and peri-urban areas), while the consumption of other root and stem vegetables has an opposite trend (although the difference between urban and peri-urban consumption is not statistically significant). People living in peri-urban areas consume less alliums compared to in urban and rural areas, while the consumption of cucurbits is lower in urban and rural compared to peri-urban areas. In all the three areas, tropical fruits claim the highest share within the total consumption of fruits (Table 9), because of their low prices compared to sub-tropical and temperate fruits (Table 10). As in case of total fruits, the consumption of individual fruits also increases as we move from the rural to urban areas, although their prices also increased from rural to urban areas. The percentage difference in fruit consumption across areas varies for various types; the difference is highest in temperate and lowest
Results and Discussion
Table 9. Per capita daily consumption of vegetables and fruits by their group and location
Vegetable/fruit Total vegetable Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Total fruits Tropical fruits Sub-tropical fruits Temperate fruits (g) Rural (%) Peri-urban (g) (%) 258.1 b 6.2 b 28.3 ab 41.4 a 24.9 a 16.0 b 129.5 a 11.9 b 69.5 b 45.4 b 19.8 b 4.3 b 100.0 2.4 11.0 16.0 9.6 6.2 50.2 4.6 100.0 65.3 28.5 6.2 (g) Urban (%) 100.0 3.6 8.7 16.7 6.6 8.0 49.2 7.2 100.0 56.9 32.5 10.7
236.7 c 6.5 b 33.1 a 28.2 b 21.4 ab 7.8 c 131.3 a 8.4 c 37.9 c 29.3 c 7.9 c 0.7 c
100.0 2.7 14.0 11.9 9.0 3.3 55.5 3.5 100.0 77.3 20.8 1.9
268.8 a 9.8 a 23.4 b 44.8 a 17.8 b 21.4 a 132.3 a 19.3 a 174.5 a 99.3 a 56.6 a 18.6 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
in tropical fruits (Table 9). The prices of fruits and vegetables, except other fruit and flower and pulses, increase as we move from rural to urban areas. The difference between rural and urban and periurban and urban areas ranges from 56–138% and 19–65%, respectively (Table 10). The difference may be attributed to structural difference in each fruit and vegetable group, value addition (such as grading, cleaning, packaging), better quality of the products itself, and partly due to the marketing cost.
Table 10. Average price of vegetables and fruits (000 VND/kg) by group and location
Vegetable/fruit Total vegetable Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Total fruit Tropical fruits Sub-tropical fruits Temperate fruits Rural 1.6 c 3.8 b 2.9 b 1.1 c 1.6 c 4.3 ab 1.5 c 7.4 a 3.5 c 2.6 b 4.3 c 6.1 a Peri-urban 2.0 b 3.9 b 3.3 b 1.5 b 2.1 b 3.8 b 1.8 b 6.0 b 5.2 b 3.9 a 6.2 b 7.5 a Urban 2.9 a 4.5 a 6.2 a 2.4 a 2.7 a 4.8 a 2.6 a 5.3 b 6.3 a 4.3 a 8.1 a 7.9 a Overall 2.2 4.1 4.1 1.8 2.1 4.3 1.9 6.1 5.3 3.8 6.6 7.8
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
An Analysis of Food Demand Patterns in Hanoi
4.4 Consumption by income group
Income is the most important factor to affect expenditure, especially expenditure on food. With improved income, people demand not only more quantity but also better quality and high dietary diversity. This is true for Hanoi also. For example, the consumption of the total quantity of food is significantly higher among upper income families compared to the consumption of families in low-income brackets. There is no significant difference in the consumption of total food consumed across families in the middle and mid-upper income groups, although these groups consume significantly more food than the households in the low-mid and low-income groups (Table 11). Table 11. Per capita daily consumption of food (g) by food and income groups
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Low 433.8 a 239.6c 41.2 d 62.6 d 31.0c 16.3 d 120.8 b 945.2 d Low-mid 410.1 b 243.8c 80.1 c 92.2 c 49.8b 19.8 c 132.9 ab 1,028.7 c Middle 408.0 b 260.8 b 97.3 b 98.7 bc 45.9b 22.6 bc 132.6 ab 1,065.9 b Mid-upper 403.3 b 250.9 bc 101.2 b 95.1 c 53.2 ab 25.9 ab 136.0 a 1,065.5 b Upper 398.3 b 276.8 a 147.0 a 121.1 a 54.8 ab 27.4 a 126.8 ab 1,152.2 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
The major difference in food consumption patterns across income groups come from higher intakes of fruits, meats, aquatic products as well as egg and milk, and lower consumption of cereals among upper income brackets. For example, the upper income group consumes 3.6 and 1.9 times more fruits and meats, respectively compared to the low-income group, and the shares of fruits and meats in consumption among the former group are 12.8% and 10.5%, respectively compared to 4.4% and 6.6% among the lowincome group (Table 11). Similar to the trends in food quantity, the daily per capita expenditure on food gradually increases from the low-income to upper income groups (Table 12). The differences in expenditure between these groups are more than proportionate differences in food quantity implying that the higher-income group spends more for the same quantity of food. The prices of different food items generally increase as we move from the low-income to the upper income groups. The difference is statistically significant across the low and upper income groups, and except egg and milk between low-mid and upper income groups. This implies that the rich usually pay higher prices for their food (Table 13), suggesting food commodities consumed by these people in each food group are of higher price or these commodities are of better quality (in terms of grade, packing, safety, etc.) than those consumed by the poor.
Results and Discussion
Table 12. Per capita daily expenditure of food (VND) by food group and income group
Food group Low Income group Low-mid Middle 1,641.8 b 495.0 c 390.3 c 2,023.6 c 582.6 b 308.5 c 698.3 c 6,140.0 c 1,733.3 a 570.1 b 534.9 b 2,290.6 b 624.5 b 379.3 b 770.5 b 6,903.1 b Mid-upper 1,658.4 b 561.8 b 612.0 b 2,260.4 b 680.9 b 447.6 ab 774.3 b 6,995.3 b Upper 1,777.8 a 660.0 a 893.8 a 2,901.4 a 825.8 a 484.5 a 878.6 a 8,421.8 a
Cereals 1,483.7 c Vegetables 408.3 d Fruits 181.7 d Meats 1,256.6 d Aquatic products 299.0 c Egg and milk 250.2 d Others 578.0 d Total 4,457.6 d
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Table 13. Average price of food (000 VND/kg) by food group and income group
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Low 3.5 d 1.8 d 4.4 b 16.9 d 11.7 c 16.9 b 6.8 c Income group Low-mid Middle Mid-upper 4.2 c 2.1 c 4.8 b 20.2 c 14.3 b 20.0 ab 7.5 b 4.5 b 2.3 b 5.5 a 21.6 b 17.0 a 19.5 ab 7.9 b 4.3 bc 2.3 b 5.6 a 21.9 ab 17.3 a 20.6 a 7.8 b Upper 4.6 a 2.4 a 5.9 a 22.8 a 19.2 a 23.2 a 9.5 a Overall 4.2 2.2 5.3 20.5 16.1 20.0 7.8
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Consumption of allium generally increases as we move from the low to upper income groups, and the difference especially between low and upper income group is statistically highly significant (Table 14). The difference in consumption of other root and stem, heading cole and cucurbits is not significant across income groups. The consumption of leafy vegetables in the upper income group is higher than all the other groups, but the other groups do not differ significantly with each other in its consumption. On the other hand, the consumption of pulses is low among the low-income group compared to all other groups, but all these other groups do not have a significant difference in their consumption. The consumption of other fruit and flower vegetables surprisingly is highest among the middle and upper income groups, and does not differ among other income groups. The consumption of all kinds of fruits increases as we move from the low to the upper income group, but the rate of increase depends upon fruit type (Table 14). The differences between upper and lower income groups in the consumption of temperate, sub-tropical and tropical fruits are 7.7, 4.1 and 3.1 times, respectively. No statistical difference in quantity of fruits consumed by the middle and the mid-upper income group
An Analysis of Food Demand Patterns in Hanoi
was observed. The shares of temperate and sub-tropical fruits consumed among upper income group are higher than those in the low-income group but an opposite is true for the tropical fruits. The expenditures on fruits and vegetables increase as one moves from low to upper income groups, except in cucurbits. The difference, however, is more frequently significant between low and upper income groups, and less frequently significant between the two adjacent income groups (Table 15). Table 14. Per capita daily consumption of vegetables and fruits (g) by their group and income group
Vegetable/fruit Low (g) (%) Low-mid (g) (%) 243.8 c 100.0 6.4 c 2.6 25.7 a 10.5 32.8 b 13.5 18.5 a 7.6 13.0 b 5.3 124.6 b 55.2 12.8 ab 5.2 80.1 c 100.0 55.6 b 69.5 18.1c 22.6 6.4 c 8.0 Middle (g) (%) 260.8 b 100.0 8.7 ab 3.3 29.7 a 11.4 40.9 ab 15.7 22.1 a 8.5 17.8 a 6.8 128.0 b 49.1 13.6 ab 5.2 97.3 b 100.0 62.4 b 64.1 27.6 b 28.3 7.3 bc 7.5 Mid-upper (g) (%) 250.9 bc 100.0 8.0 b 3.2 27.7 a 11.0 36.2 ab 14.4 22.2 a 8.8 13.8 b 5.5 128.5 b 51.2 14.6 a 5.8 101.2 b 100.0 60.4 b 59.7 30.5 b 30.2 10.3 ab 10.1 Upper (g) (%) 276.8 a 100.0 9.9 a 3.6 32.1 a 11.6 43.0 a 15.5 20.7 a 7.5 17.2 a 6.2 140.1 a 50.6 13.9 ab 5.0 147.0 a 100.0 83.2 a 56.6 50.8 a 34.6 13.1 a 8.9
Total vegetable 239.6 c 100.0 Allium 4.8 d 2.0 Other root and stem 28.1 a 11.7 Heading cole 35.4 ab 14.8 Cucurbits 22.8 a 9.5 Other fruit and flower 12.0 b 5.0 Leafy 126.4 b 52.8 Pulses 10.1 b 4.2 Total fruit 41.2 d 100.0 Tropical fruits 27.2 c 66.1 Sub-tropical fruits 12.3 d 29.9 Temperate fruits 1.7 d 4.1
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Table 15. Per capita daily expenditure on vegetables and fruits (VND) by income groups
Vegetable/fruit group Allium Other root and stem Heading cole Cucurbits Other fruits and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Low 12.9 d 48.7 b 43.2 b 42.1 a 32.7 c 183.7 c 45.0 c 93.0 d 79.4 d 9.3 c Income group Low-mid Middle 20.0 c 66.9 a 49.2 b 35.0 a 37.9 bc 238.2 b 47.8 bc 193.4 c 147.1 c 49.8 b 27.2 b 72.7 a 67.2 a 44.7 a 57.9 a 240.2 b 60.1 ab 242.1 bc 243.9 b 49.0 b Mid-upper 30.9 ab 65.4 a 70.8 a 46.0 a 41.9 b 244.6 b 62.2 ab 251.0 b 278.9 b 82.1 a Upper 35.1 a 84.9 a 79.0 a 49.3 a 58.2 a 284.3 a 69.3 a 353.2 a 448.0 a 92.5 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Part of the difference in expenditures on fruits and vegetables across various groups is explained in terms of quantity consumed by these groups, and remaining in terms of the quality of these products reflected in the difference of prices across these groups.
Results and Discussion
Consistent with expenditure trends, the prices of fruits and vegetables increase as one moves from low to upper income groups, although the difference is more significant across the distant income groups, rather than adjacent groups (Table 16). Table 16. Average prices of fruits and vegetables (000 VND/kg) by income groups
Vegetable/fruit group Allium Other root and stem Heading cole Cucurbits Other fruits and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Low 3.9 a 3.0 b 1.3 c 2.0 bc 3.9 b 1.6 d 6.5 ab 3.2 b 5.2 c 6.4 c Income group Low-mid Middle 4.2 a 4.3 a 1.6 b 1.9 c 4.1 ab 1.8 c 5.2 c 3.6 ab 6.0 b 7.8 ab 4.0 a 5.2 a 1.8 b 2.2 abc 4.5 ab 2.0 b 6.4 ab 3.9 a 6.7 a 8.4 ab Mid-upper 4.2 a 4.1 a 2.1a 2.2 ab 4.4 ab 2.1 a 5.6 bc 3.8 a 7.2 a 8.4 a Upper 4.1 a 3.7 a 2.0 ab 2.4 b 4.8 a 2.1 a 6.8 a 4.2 a 7.4 a 7.0 bc
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
4.5 Consumption by farm type
The consumption of total food was the highest among non-farmer urban consumers, while it was the lowest among non-farmers in rural areas (Table 17). This is true for almost all individual foods, except that the non-farmer urban group consumes the lowest amount of cereals. Vegetable farmers consume significantly more food than non-vegetable farmers and non-farmers in rural areas, but their consumption is similar to the non-farmers in peri-urban areas, but lower than non-farmer urban consumers. Cereal consumption was highest among farmers, and vegetable farmers consume slightly more cereals perhaps because they have to do more manual work in harvesting Table 17. Per capital daily consumption of food (g) by food group and farm type
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Farmer type Vegetable Non-vegetable 456.4 a 253.8 b 36.4 d 69.3 d 47.9 ab 17.4 c 142.8 a 1,024.1 b 441.7 b 239.9 c 45.6 c 80.6 c 42.7 b 15.3 cd 129.5 b 995.3 c Rural 397.9 c 214.2 d 55.4 bc 75.7 cd 31.7 c 11.2 d 134.9 ab 920.9 d Non-farmer type Peri-urban 383.5 c 263.6 ab 97.7 b 101.9 b 46.3 ab 23.8 b 129.7 ab 1,046.5 b Urban 360.9 d 269.2 a 177.6 a 117.3 a 50.7 a 33.4 a 122.0 b 1,131.2 a
Mean separation in columns by Duncan’s multiple range test, P < 0.10.
An Analysis of Food Demand Patterns in Hanoi
of vegetables. Among non-farmers, people in peri-urban and rural areas consume more rice than urban dwellers. Non-farmers in urban and peri-urban areas consume the highest amounts of vegetables followed by vegetable farmers. The vegetable farmers consume more vegetables and other foods, but lower amounts of fruit and meat compare to non-vegetable farmers. Non-farmers in peri-urban areas consume significantly less food in total compared to non-farmers in urban areas, although a mixed trend is observed for individual food items (Table 17). The average per capita daily expenditure on food items of non-farmers in urban areas is 33% and 108% higher than non-farmers in peri-urban and rural areas, respectively (Table 18). This partly reflects more preferences for high value commodities such as meat, fruit, etc. and partly because of higher food prices attributed to value addition as it travels from rural to urban areas. Although vegetable farmers consume more food than non-vegetable farmers, per capita food expenditure is about the same on both the group of farms. Table 18. Per capital daily expenditure of food (VND) by food group and farm type
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Farmer type Vegetable Non-vegetable 1,502.5 c 430.6 c 135.8 d 1,329.9d 442.6 c 278.7 c 670.9 c 4,790.9 c 1,455.3 cd 397.6 d 180.3 c 1,556.5c 397.5 c 233.5 c 624.8 c 4,845.4 c Rural 1,413.6 d 343.1 e 167.0 cd 1,374.8 cd 295.8 c 193.6 c 743.9 bc 4,531.7 c Non-farmer type Peri urban 1,765.7 b 588.2 b 520.1 b 2,409.3 b 636.4 b 385.4 b 763.9 b 7,069.0 b Urban 1,956.9 a 753.7 a 1,141.8 a 3,160.2 a 914.1 a 596.4 a 889.0 a 9,412.0 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Non-farmers in urban and peri-urban areas consumed the highest amounts of all individual vegetable and fruit items, with the exceptions of root and stem as well as leafy vegetables (Table 19). Between these two non-farmer groups, non-farmers in urban areas consumed significantly higher amounts of allium, other fruits and flower, pulses, and all different types of fruits. Non-farmers in rural areas consumed the lowest amounts of almost all different types of vegetables and fruits; however, the differences across non-farmer and farmer groups in rural areas were not statistically significant in most cases. Although vegetable farmers consumed significantly higher amounts of total vegetables compared to non-vegetable farmers, for individual vegetables the only significant difference was for pulses.
Results and Discussion
Table 19. Per capita daily consumption and percent share of vegetables and fruits by their group and farm type
Vegetable/fruit Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Farmer type Vegetable Non-vegetable (g) (%) (g) (%) 6.1 c 37.0 a 32.4 b 19.7 a 11.9 c 134.1 a 12.6 b 27.4 d 9.2 c 1.2 c Rural (g) (%) Non-farmer type Peri-urban (g) (%) Urban (g) (%)
2.4 5.9 c 2.5 6.0 bc 2.8 8.1b 3.1 10.0 a 3.7 14.6 30.0 ab 12.5 29.7 ab 13.9 24.8 c 9.4 23.5 c 8.7 12.8 30.6 b 12.8 27.7 b 12.9 48.4 a 18.4 45.5 a 16.9 7.8 24.9 a 10.4 10.4 b 4.9 23.1 a 8.8 17.8 a 6.6 4.7 10.0 c 4.2 9.7 c 4.5 17.5 b 6.6 21.3 a 7.9 52.8 129.9 a 54.1 119.9 a 56.0 130.7 a 49.6 131.8 a 48.9 5.0 8.6 c 3.6 10.7 bc 5.0 11.1 bc 4.2 19.4 a 7.2 72.7 36.4 cd 74.3 61.1 c 86.9 63.5 b 64.2 102.6 a 57.2 24.4 10.7 c 21.8 9.2 c 13.1 28.8 b 29.1 57.9 a 32.3 3.2 1.9 c 3.9 0.0 d 0.0 6.7 b 6.8 18.9 a 10.5
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Fruit consumption was highest in the non-farmer group in urban areas and surprisingly was the lowest among vegetable farmers. Perhaps vegetable farmers compensate their relatively low fruit consumption with higher consumption of vegetables. Tropical fruits are consumed more than other fruits in the daily diet in the sample area. The expenditure on fruits and vegetables of non-farmers in urban areas is highest among different groups. This is true for almost all individual fruit and vegetable items (Table 20). Higher prices for individual fruits and vegetables in addition to higher consumption of these commodities in urban areas, explain these results. For example, non-farmers living in urban had to pay 2,500 VND/kg for leafy vegetables, while vegTable 20. Per capital daily expenditure (VND) on fruit and vegetable group by farm type
Vegetable/fruit Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Farmer type Vegetable Non-vegetable 17.3 c 70.1 b 37.6 c 33.5 c 27.7 c 185.1 c 59.3 b 76.8 d 53.7 c 6.9 c 15.4 c 52.9 cd 34.9 cd 42.8 bc 26.7 c 181.9 c 43.3 c 111.6 cd 66.6 c 11.4 c Rural 12.7 c 44.7 d 29.5 d 15.9 d 24.4 c 173.4 c 42.4 c 141.7 c 53.1 c - Non-farmer type Peri-urban 25.0 b 64.6 bc 79.4 b 54.9 a 57.3 b 259.1 b 47.0 bc 238.9 b 235.0 b 50.9 b Urban 41.3 a 89.7 a 102.4 a 46.6 ab 75.1 a 326.1 a 76.6 a 458.3 a 547.5 a 154.8 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
0
An Analysis of Food Demand Patterns in Hanoi
etable farmers and both non-vegetable farmers and non-farmers in rural areas paid only about 1,400 VND/kg (Table 21). Other trends in expenditure are similar to those in the quantities. Table 21. Average price of vegetable and fruit groups (000 VND/kg) by farm type
Vegetable/fruit Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Farmer Vegetable Non-vegetable 2.8 1.9 1.2 1.7 2.3 1.4 4.7 2.8 5.8 5.8 2.6 1.8 1.1 1.7 2.7 1.4 5.0 3.1 6.2 6.0 Rural 2.1 1.5 1.1 1.5 2.5 1.4 4.0 2.3 5.8 - Non-farmer Peri-urban 3.1 2.6 1.6 2.4 3.3 2.0 4.2 3.8 8.2 7.6 Urban 4.1 3.8 2.3 2.6 3.5 2.5 3.9 4.5 9.5 8.2
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
4.6 Consumption by season
The overall food consumption is lowest during the cold-wet season and highest during the hot-wet season; however, the consumption during cold-dry season is not significantly different compare to consumption during the other two seasons. A significant variation in the consumption of individual food items can be noticed across the seasons. For example, vegetable consumption is lowest during the hot-wet and highest during the cold-dry season, while fruit consumption is just the reverse. The cereal consumption is also low during the hot-wet season, while the consumption of egg and milk, and other food are highest during the hot-wet season. So shortage of vegetables and to some extent cereals during the hot-wet season is compensated by the abundant availability of fruits, egg and milk and other foods in the season. No significant difference in the consumption of meats and aquatic products was observed across the three seasons (Table 22). Table 22. Per capita daily consumption of food (g) by food group and season
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Cold-wet 415.3 a 242.5 b 84.3 b 91.5 a 46.9 a 19.0 c 120.3 b 1,019.6 b Season Cold-dry 419.6 a 289.4 a 66.0 c 88.0 a 42.4 a 21.8 b 123.5 ab 1,050.8 ab Hot-wet 401.6 b 228.4 c 128.5 a 97.3 a 49.3 a 25.4 a 136.2 a 1,066.7 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Results and Discussion
Some of these differences in consumption are explained by the difference in prices across season. For example, highest prices of vegetables during the hot-wet season and highest prices of fruits during the cold-dry season explain their lowest consumption in the respective season (Table 23). Table 23. Average price of food (000 VND/kg) by food group and season
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Cold-wet 3.8 2.1 4.7 22.1 11.6 18.1 5.9 Season Cold-dry 4.2 1.8 6.3 23.4 12.2 16.2 5.5 Hot-wet 4.1 2.4 5.6 22.9 14.3 16.0 5.8 Overall 4.0 2.1 5.5 22.8 12.7 16.6 5.7
Differences in prices and consumption explain the variation in food expenditures across season. The hot-wet season accounted for the highest share (36%) of annual food expenditures, while the cold-wet season had the lowest share (31%) (Fig. 1). The highest consumption and expenditure on overall food during the hot-wet season is mainly because of high consumption of fruits, especially tropical, during this season.
Hot-wet 36%
Cold-wet 31%
Cold-dry 33%
Fig. 1. Seasonal share in annual expenditure of overall food Despite a significant difference in consumption, expenditures on vegetables hardly varied across season. On the other hand, although there was no significant difference in the consumption of cereals across season, a significantly higher expenditure was made during the cold-dry season (Table 24). This suggests that consumers maintained their level of cereal consumption irrespective of its price, while they adjusted their consumption levels of vegetables according to fluctuating prices. The highest expenditure on fruits during the hot-wet season reflects consumers’ preferences for fruits although their prices
An Analysis of Food Demand Patterns in Hanoi
were not at the lowest level during this season. The high supply of fruits from domestic production as well as an attempt of the household to mitigate the shortage of vegetables by substituting for fruit during this season may explain this preference. Table 24. Per capita daily expenditure of food (VND) by food group and season
Vegetable/fruit Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Cold-wet 1,581.2 c 508.6 a 398.4 b 2,022.7 b 542.1 b 343.2 b 707.3 a 6,103.5 b Cold-dry 1,761.6 a 535.3 a 418.9 b 2,061.1 b 518.1 b 352.7 b 681.5 a 6,329.1 b Hot-wet 1,633.5 b 557.5 a 725.5 a 2,224.4 a 703.9 a 406.8 a 784.5 a 7,036.2 a
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Leafy vegetables remain as the major vegetable group across all seasons, although its share in the diet decreases from 70% during the hot-wet season to 37% during the cold-dry season. Cucurbit consumption is also highest during the hot-wet season. All other vegetables, except pulses and vegetables in total, are consumed in highest quantities during the cold-dry season (Table 25). This is because the cold-dry season is the main season of vegetable production in northern Vietnam (Thuy et al. 2002); therefore, the greater availability of vegetables during this season makes most of these individual vegetables less expensive compared to during the hot-wet and cold-wet seasons. For example, heading cole costs just 1200 VND/kg in the cold-dry season but costs up to 4700 VND/kg in the hot-wet season (Table 26). Table 25. Per capita daily consumption (g) of vegetable and fruits group by season
Vegetable/fruit Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits (g) Cold-wet Share (%) 3.1 12.7 16.6 5.2 5.5 51.5 5.4 53.5 27.8 18.7 (g) Cold-dry Share (%) 2.8 16.4 23.3 7.4 8.3 37.4 4.4 47.0 45.0 8.0 (g) Hot-wet Share (%) 2.8 3.3 2.2 13.0 2.9 70.0 5.7 77.8 21.3 0.9
7.6 a 30.7 b 40.2 b 12.7 c 13.4 b 124.9 b 13.0 a 45.1 b 23.4 b 15.8 a
8.2 a 47.6 a 67.3 a 21.5 b 23.9 a 108.3 c 12.6 b 31.0 c 29.7 a 5.3 b
6.5 b 7.5 c 5.1 c 29.8 a 6.7 c 159.8 a 13.0 a 100.0 a 27.4 ab 1.1 c
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
Results and Discussion
Table 26. Average price of vegetables and fruits (000 VND/kg) by season
Vegetable/fruit Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Cold-wet 3.0 2.7 2.0 2.0 3.0 1.6 3.9 2.7 6.9 7.3 Cold-dry 3.1 1.8 1.2 2.0 2.3 1.8 3.7 4.2 8.2 8.3 Hot-wet 3.9 4.6 4.7 2.1 5.8 1.9 5.5 4.3 10.3 14.4
The consumption of the tropical fruits in the hot-wet season was 3.2 and 2.2 times higher than in the cold-dry and cold-wet seasons, respectively (Table 25). An attempt of consumers to substitute fruit consumption for the seasonally low vegetable supply as well as to meet the high water requirement of the human body in the hot-wet season explain these results, although the prices of tropical fruits are not lowest during this season (Table 26). Individual vegetables showed different expenditure patterns during different seasons. The per capita expenditure was highest on heading cole and other stem and root vegetables during the cold-dry and cold-wet seasons; on cucurbits, leafy, and pulses during the hot-wet season; and on other fruit and flower vegetables during cold-dry season. No significant differences were observed across season for allium (Table 27). Expenditures on tropical and sub-tropical fruits are highest during the hot-wet season (Table 27), although their prices are highest during this season (Table 26). Again this reflects an effort of consumers to substitute the high price of vegetables with fruits to meet the high water requirement during this season. The highest expenditure for temperate fruits was during the cold-wet season. Table 27. Per capita daily expenditure (VND) on vegetable and fruit groups by season
Vegetable/fruit Allium Other root and stem Heading cole Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Sub-tropical fruits Temperate fruits Cold-wet 23.0 a 82.5 a 80.7 a 24.9 c 40.7 b 205.7 b 51.3 b 122.2 b 161.0 c 115.1 a Cold-dry 25.4 a 87.2 a 78.5 a 42.6 b 56.1 a 199.0 b 46.5 b 129.9 b 244.8 b 44.2 b Hot-wet 25.5 a 34.8 b 23.9 b 62.2 a 38.7 b 300.7 a 71.6 a 426.9 a 282.9 a 15.8 c
Mean separation in rows by Duncan’s multiple range test, P < 0.10.
An Analysis of Food Demand Patterns in Hanoi
4.7 Sources of food
4.7.1 Overall Food comes from various sources but temporary markets, retail markets, and ownedfarms were the major sources. Street vendors, home gardens and gifts also supplied 2–6% of the total food requirements, while night markets, supermarkets, and vegetable shops provided insignificant food quantities (less than 1%) (Table 28). Surprisingly, most individual food items (except cereals) in and around Hanoi were bought from temporary markets. Figuié (2003) also speculated that temporary street markets were the main source of fresh vegetable supply. The temporary market in this study was defined as a place that was not belonging to the market system established by the authorities, neither was recognized by the authorities as a market place, nor was it permanent. Most of goods sold here were foods and it was usually located at a convenient place (for example, a roadside). Prices of foods in the temporary market were usually lower than in other markets because the shop owners do not have to pay any market fees (although they may have to pay bribes to local police). Table 28. Source of food (% of quantity) by group
Temporary Food group Owned Home farm 48.65 6.46 1.60 1.77 9.89 1.40 1.42 Street Retail Night Vegetable Super shops 0.00 0.04 0.03 0.04 0.02 0.12 0.06 market 0.00 0.09 0.04 0.02 0.00 0.00 0.02 Gift 0.30 3.65 4.22 0.61 4.11 0.38 0.24 market 26.31 39.62 46.10 48.72 40.05 50.26 53.55 garden vendors market market 0.16 7.57 5.17 3.29 0.93 10.16 1.27 4.09 6.98 9.18 5.29 7.87 3.06 6.48 20.48 35.58 33.67 40.26 37.13 34.62 36.96 0.00 0.01 0.00 0.00 0.00 0.00 0.00
Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others
About one-half of the cereal supply comes from owned farms, while other foods mostly come from temporary markets and retail markets. Street vendors are the third most important source of fruits, meats, and others. The home garden is the fourth most important food source in most items, while its contribution in supplying vegetables and egg and milk (8% and 10%, respectively) surpasses that of street vendors. About 3–4% of fruits, vegetables and aquatic products were traded as gifts among consumers. Supermarkets supply less than 0.1% of vegetables and 0.05% of fruits. Tan Loc (2002) estimated less than 2% of all fresh vegetables come from stores and supermarkets. Vegetable shops, considered to be providing quality vegetables, such as pesticide residue-free produce, supply an insignificant proportion of the total vegetable supply. They also provide miscellaneous food items (although in small proportion) besides vegetables and fruits.
Results and Discussion
4.7.2 Region The relative importance of various food sources varies significantly across the three areas (Table 29). The quantity of foods bought from temporary markets, retail markets and vegetable shops tends to increase from rural to urban areas while an opposite trend was observed for owned-farm, home garden, and gifts. In the rural area, the importance of home gardens is higher compared to farm production in supplying vegetables, fruits, meats, and egg and milk; however, the farm is a more important source in supplying cereals, aquatic products and others. Twenty-five percent of eggs and milk (mostly eggs), 20% of fruits, and a 14% of vegetables consumed in rural areas came from home gardens. Even in the peri-urban area, the importance of home garden production is greater than farm production in supplying fruits, meats, egg and milk and “others”, while both Table 29. Food sources (% of quantity) by region and food group
Food group Urban Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Peri-urban Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Rural Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Owned Home Street Retail Night Vegetable Super farm garden vendors market market shops market 0.48 0.78 0.01 0.68 0.00 0.18 0.02 0.42 44.51 8.98 0.41 0.13 5.34 0.00 0.97 20.79 80.74 9.20 8.52 4.44 20.86 4.69 2.91 40.88 0.01 0.07 0.00 0.00 0.00 0.55 0.00 0.03 0.01 7.47 6.85 2.88 2.62 9.32 2.75 3.19 0.37 14.16 20.09 7.41 0.73 25.34 0.86 5.44 7.62 9.72 10.91 1.01 0.83 3.05 1.54 6.82 3.63 0.39 3.59 0.33 0.15 0.77 0.36 1.92 2.31 10.32 10.97 14.44 18.27 5.47 15.89 8.08 38.14 39.90 35.55 40.76 45.88 35.17 37.10 38.61 25.39 42.91 35.48 47.82 42.09 44.93 44.80 35.92 6.18 25.50 24.97 33.19 26.87 22.96 29.62 17.91 0.00 0.04 0.00 0.00 0.00 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.14 0.05 0.00 0.07 0.29 0.00 0.05 0.01 0.00 0.00 0.13 0.00 0.00 0.16 0.04 0.00 0.00 0.00 0.00 0.00 0.00 0.002 0.0003 0.00 0.27 0.06 0.06 0.00 0.00 0.01 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.05 0.01 Temporary Gift market 0.33 0.31 1.92 0.31 2.58 0.27 0.13 0.66 0.25 4.38 5.51 0.58 2.92 0.37 0.08 1.72 0.31 5.92 10.13 0.99 6.09 0.57 0.47 2.40 53.42 48.78 51.50 57.18 50.64 60.50 61.20 53.32 26.19 35.86 48.15 48.13 46.88 44.61 50.87 36.41 10.09 34.90 25.32 39.54 27.18 40.97 50.20 25.28
An Analysis of Food Demand Patterns in Hanoi
have similar contributions in supplying vegetables. Home gardens and farms are insignificant sources of food supplies in the urban area. Surprisingly, street vendors are more important food sources in rural area than in urban and peri-urban areas, except in supplying fruits and vegetables, where their contribution in the urban area is similar to that in the rural area. Overall, more than one-half of food consumed in the urban area is bought from temporary markets and one-third from retail markets. The quantity of foods that comes from temporary markets and retail markets are similar in peri-urban at around 36%, while only one-fourth of the total food comes from this source in rural areas. The supply of individual food groups from this source is around the average value for all food in urban areas, while in peri-urban and rural areas cereals from this source is far less than the average value and vice versa is true for other food groups. 4.7.3 Income group3 Overall, temporary markets and retail markets are the two major sources of food among the middle and upper income groups while owned farm is the major source for the low income group. The percentage of food obtained from temporary markets, retail markets, street vendors and vegetable shops generally increases as we move from lower to upper income groups, but that of owned-farms, home gardens and gifts show an opposite trend. This is true for overall as well as for individual food items. The share of food from vegetable shops is insignificant for all groups, but increases as we move from lower to upper income groups (Table 30). Own-farm was the most important source of cereal supply in all of the three income groups, and temporary and retail markets were respectively second and third sources in the order of priority for all income groups. The importance of the own-farm supply for other food items, however, varied across the income groups. For example, for the low-income group, it was the third important source for vegetables and aquatic products supplies after temporary market and retail market. For the middle-income group, ownedfarm supplies for aquatic products still ranked at the third position, while for vegetables it is ranked at fifth after temporary markets, retail markets, home gardens, and street vendors in that order. The importance of owned-farm supplies decreased even further for the upper-income group in supplying these two items. The home garden was the third important source for fruits, meats and egg and milk at the low-income group. For the middle and upper income groups, its importance remained at the third position for egg and milk while street vendors took over for the fruits and meats.
3
As no statistical difference in the quantity and expenditure of food between low and low-mid, and between middle and mid-upper income groups was observed in the earlier analysis, the following analysis has merged the former two into the low income group, and later two into the middle income group.
Results and Discussion
Table 30. Food sources (% of quantity) by income and food groups
Income group/ food group Low Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Middle Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Upper Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Owned Home Street Retail Night Vegetable Super- Temporary farm garden vendors market market shops market Gift market 63.73 12.34 1.96 2.01 9.31 1.84 1.72 33.21 44.67 5.20 1.40 1.28 12.00 1.19 1.45 19.45 37.76 2.64 1.95 3.01 3.01 1.65 0.89 14.43 0.18 9.79 15.08 5.88 3.72 20.64 2.16 4.32 0.20 7.64 4.33 2.31 0.06 8.33 1.18 2.80 0.00 4.27 2.83 4.13 1.44 5.41 0.29 2.04 2.50 3.89 5.01 3.74 5.91 1.27 6.01 3.59 4.51 7.47 9.58 4.81 8.66 4.25 6.30 6.13 5.26 9.57 10.02 7.86 6.86 1.01 7.86 7.46 16.11 36.33 32.11 40.99 43.63 38.58 42.16 28.18 21.49 34.67 34.05 40.58 34.68 33.39 35.64 30.19 24.22 37.67 33.41 38.77 39.77 34.56 34.31 32.29 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.00 0.00 0.00 0.005 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.04 0.00 0.00 0.00 0.10 0.00 0.01 0.005 0.03 0.00 0.06 0.04 0.17 0.00 0.02 0.00 0.10 0.12 0.00 0.00 0.00 0.35 0.08 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.12 0.06 0.00 0.00 0.00 0.03 0.04 0.00 0.08 0.00 0.11 0.00 0.00 0.01 0.03 0.17 7.08 6.77 1.16 8.92 0.19 0.13 2.54 0.33 2.98 4.28 0.57 3.39 0.53 0.22 1.48 0.40 1.21 2.92 0.30 2.25 0.11 0.50 0.99 17.31 30.53 39.07 46.22 28.50 37.38 47.82 28.14 28.81 41.86 46.30 50.40 41.17 52.14 55.17 39.89 32.37 44.46 48.74 45.82 46.67 57.26 55.79 42.67
4.8
Farming-based group
About 38% of the food consumed among farm families came from their own-farm production, and an additional 5% came from home gardens. About 76% of cereals, 18% of aquatic foods, and 11% vegetables came from own-farm production, while 21% of egg and milk, 15% of fruits, and 12% of vegetables consumed by farm families came from their home gardens. Temporary and retail markets were the second and third most important sources of food for farmers, respectively, surpassed only by own-farm production. A significant percentage of aquatic products, other food, and meats were purchased from street vendors by farmers. Surprisingly, more than one-tenth of the fruits were shared as gifts among farm families (Table 31). Among non-farm families, temporary markets were the single major source of food purchases followed by retail markets. About 6% of the food purchases came from street vendors, and 1% from their home gardens (Table 31).
An Analysis of Food Demand Patterns in Hanoi
Table 31. Food sources (% of quantity) by farming based and food groups
Farming base/ Temporary food group Farmer Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Non-farmer Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Owned Home farm 75.91 11.34 4.87 2.98 18.01 2.90 2.56 38.13 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Street Retail Night Vegetable Super- Gift 0.12 5.76 10.62 0.77 5.79 0.50 0.27 2.24 0.61 0.91 1.83 0.46 2.07 0.28 0.23 market 11.51 33.11 35.24 41.46 30.72 41.81 48.95 26.41 52.23 48.04 51.45 55.80 51.28 57.33 58.96 garden vendors market market shops market 0.18 12.08 15.25 6.54 1.69 20.50 2.11 4.80 0.89 1.80 1.24 0.68 0.00 1.55 0.37 2.21 6.50 5.20 9.58 13.57 3.38 9.98 5.51 7.42 7.54 10.13 1.07 0.91 2.79 2.10 10.08 31.21 28.82 38.60 30.22 30.92 36.11 22.90 38.85 41.39 35.25 41.95 45.69 37.82 38.16 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.03 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.07 0.00 0.00 0.02 0.01 0.00 0.10 0.04 0.00 0.05 0.23 0.14 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.20 0.05 0.05 0.00 0.00 0.05
4.9 Food quality
Not much quantitative research has been conducted on the status of food quality in Hanoi. Bridier (2000) estimated consumers’ perception of vegetable quality in the city. The following qualitative measures of “good vegetable” were listed in the study: 1. Desirable appearance, freshness, taste, tenderness, and size 2. Clean, healthy, good nutritional quality, locally grown during the main season, grown in a healthy environment with a reasonable use of agrochemical inputs4 3. Well preserved and well packed 4. Good price Figuié (2003) quantified the consumers’ perception about safe food in Hanoi. A large number of respondents listed vegetables (88.5%), meat (69.5%), and fruits (46.0%) as the three most dangerous foods for consumers’ health. Kangkong, brassicas, and yardlong bean were specifically considered dangerous by a large number of respondents. About 81% of respondents expressed their concerns about the high use of agrochemicals in producing and preserving vegetables, meat, fruits, and fish. However, about half of the consumers thought that they can more or less protect themselves against these risks, while the other half thought that they could only partially protect themselves. Consequently, the majority of the interviewees (89%) believed that consuming foods they have prepared
4
Some consumers, however, looked for vegetables with imperfection, even with insects, to satisfy their impression of good vegetables with minimal spray of pesticides.
Results and Discussion
“does not present” or “hardly present” a health risk. Therefore, these consumers may not be willing to pay additional prices for low agro-chemical residues on foods. Choosing, cleaning and cooking food were considered to be essential links of the quality chain. In the present study, we consider five criteria of good quality food: 1. Supply of nutrients, including micronutrients 2. Food diversity 3. If the food is processed or not 4. Proportion of the food eaten outside 5. Difference in prices In the following sections, we compare food quality across different groups with respect to these criteria. 4.9.1 Nutrient availability and source Overall. When overall mean values of the sample were compared with the midpoint of the requirement ranges, dietary deficiencies in calcium, vitamins B1, B2 and niacin were identified. When nutrient availability was compared on daily basis, a significant population fell below 80% of dietary requirements for almost all nutrients, reflecting imbalanced diets and poor food quality. For example, 67%, 81%, 95%, and 75% population was below the 80% of the requirements for calcium, and vitamins B1, B2, and niacin, respectively. A smaller percentage of the population was deficient in calories, iron, and vitamins A and C as well (Table 32).
Table 32. Daily per capita availability and deficiency level of major and micronutrients
Nutrient (%)1 Calories Nonfat Fat Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C
1
Unit kcal kcal kcal g mg mg IU mg mg mg mg
Recommended level1 1,800–2,400 - - 45-65 800-1,200 10-15 4,200-5,000 1.12 1.22 14.66 50-70
Nutrient availability 2,226.7 1,766.8 459.9 94.5 705.1 19.8 6,365.7 0.7 0.5 10.3 87.3
Population > 20% deficit 17.0 6.9 67.3 14.5 30.4 80.6 94.9 75.3 24.2
The recommended level of nutrient consumption was taken from Food and Nutrition Board (1989), and the deficiencies
0
An Analysis of Food Demand Patterns in Hanoi
Despite an increasing food diversity of the diet of Hanoi and its surrounding people, cereals are the main sources of calories, protein, and vitamins B1 and niacin. About one-fifth of the calories consumed in Hanoi and its surrounding areas were fat-based, and the remaining four-fifths came from nonfat sources. Vegetables provided more than three-fourths of the vitamins A and C in diets, and were the major source of calcium. Egg and milk were only minor sources of calcium. Similarly, iron came from non-traditional sources such as cereals and “others”, which have low bioavailability, rather than from meats, which have high bioavailability. Most of the nonfat-based calories came from cereals, and most of the fat-based calories came from meats (Table 33). Table 33. Nutrient source by food group
Food group Unit Calories Nonfat Fat Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C kcal kcal kcal g mg mg IU mg mg mg mg Percentage supplied by Recommended Total Aquatic Egg and level nutrients Cereals Vegetables Fruits Meats products milk Others 1,800–2,400 - - 45–65 800–1,200 10–15 4,200–5,000 1.12 1.22 14.66 50–70 2,226.7 1,766.8 459.9 94.5 705.1 19.8 6,365.7 0.7 0.5 10.3 87.3 60.6 74.4 7.6 36.2 17.9 27.8 0.0 54.4 29.1 64.5 0.8 3.4 3.8 1.7 6.7 23.5 15.3 76.5 27.7 36.3 16.9 79.1 1.3 1.5 0.5 0.6 2.7 3.5 14.9 4.1 5.3 2.1 18.6 15.6 5.4 54.6 12.2 1.1 4.8 1.7 7.2 7.8 9.5 0.7 1.6 1.4 2.2 6.4 12.8 0.9 0.0 0.3 2.2 2.1 0.0 2.8 2.3 4.6 2.5 2.8 2.3 5.0 3.6 11.5 0.3 0.1 14.8 11.2 28.8 35.3 39.1 45.3 1.9 2.7 7.8 4.6 0.7
Income group. The nutrient availability of almost all major micronutrients steadily increased as we moved from low to upper income groups, suggesting some improvement in food quality with respect to nutrient supply at higher income levels. The difference between the low and upper income groups was significant for all nutrients, except iron. However, it was not significant between the middle and upper income groups for calories, protein, calcium, and iron. There was no significant difference in nonfat-based calories across income groups, but fat-based calories and their share in total calories increased as one moved from low to upper income groups (Table 34). In supplying vitamin A, vegetables played a higher role among the low income group while fruits were more important among the high income group; similarly relatively low bioavailable iron from cereals were more important among the low income group and relatively high bioavailable iron from meats were a more important source among the high income group (Appendix 4). Generally, the population deficient in meeting at least 80% of its nutrient requirements increased as income decreased. However, a significant proportion of the population remained deficient (even among the upper income group) especially for the nutrient deficient at the mean level, suggesting that lack of income is only one factor. A total
Results and Discussion
Table 34. Daily per capita availability and deficiency level of nutrients
(%)1 Nutrient Calories Non-fat Fat Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C
1
Recommended Unit kcal kcal kcal g mg mg IU mg mg mg mg level1 1,800–2,400 - - 45–65 800–1,200 10–15 4,200–5,000 1.12 1.22 14.66 50–70 Low 2,183.7 b 1,761.3 a 422.3 c 91.3 b 668.9 b 19.6 a 5,789.1 c 0.71 c 0.48 c 9.9 c 76.0c
Nutrient availability Middle 2,250.2 a 1,770.1 a 480.1 b 97.0 a 719.5 a 20.2 a 6,511.6 b 0.74 b 0.54 b 10.3 b 90.8 b Upper 2,276.1 a 1,772.3 a 503.8 a 96.3 a 760.0 a 19.5 a 7,435.9 a 0.78 a 0.61 a 11.2 a 106.9 a
Population > 20% deficit Low 20.3 - - 10.2 68.3 18.0 33.3 84.4 96.6 79.6 30.9 Middle 14.6 - - 5.2 66.2 12.8 28.3 79.1 94.7 74.5 21.2 Upper 14.5 3.0 67.3 9.6 28.5 74.9 90.9 66.8 15.2
The recommended level of nutrient consumption was taken from Food and Nutrition Board (1989), and the deficiencies
of 67%, 75%, 91%, and 67% of the upper-income group consumed less than 80% of its requirements for calcium, and vitamins B1, B2 and niacin, respectively in this group (Table 34). This suggests that resources to overcome micronutrient deficiency are not the major cause; rather lack of knowledge about the importance of nutrients in their diet produced these results. Region. The availability of calcium and vitamins B1 and B2 increased as we moved from rural to peri-urban to urban areas reflecting some improvement in food quality in terms of nutrient supply. The difference in the availability of protein was not statistically different across the urban and peri-urban regions, but its consumption was the lowest in rural areas. The consumption of vitamins A and C were statistically similar across peri-urban and rural areas, and were highest in urban areas. There were no significant differences in the consumption of vitamins B3 across regions, suggesting no difference in food quality across the regions in terms of supplying this micronutrient. Consumption of calories consumption was highest in peri-urban areas. Consumption of iron was lowest in rural but highest in peri-urban areas. The consumption of fat-based calories increased from rural to peri-urban to urban areas; while non-fat based calories consumption was highest in peri-urban and rural areas (Table 35). General speaking, the percentage of population failing to meet 80% of their dietary requirements for nutrients was highest in rural areas and lowest in urban areas, which reflects an improvement in food quality as we move toward the urban center. The population deficient in calories was highest in rural areas, followed by urban and then peri-urban areas (Table 35). A high portion of the population (even in urban areas) did not meet 80% of its requirements for calcium and vitamins B1, B2 and B3. This indicates widespread unawareness about the importance of micronutrients as quality elements in their diet. Moreover, average income level data may partially conceal the widespread presence of poverty in urban areas, which is revealed in nutrient deficient populations
An Analysis of Food Demand Patterns in Hanoi
Table 35. Daily per capita availability and deficiency level of nutrients by location
(%)1 Nutrient Calories Nonfat Fat Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C
1
Recommended Unit kcal kcal kcal g mg mg IU mg mg) mg mg level1 1,800–2,400 - - 45–65 800–1,200 10–15 4,200–5,000 1.12 1.22 14.66 50–70 Urban 2,181.3 b 1,659.1 b 522.2 a 99.4 a 776.6 a 20.2 b 7,499.3 a 0.8 a 0.61a 10.4 a 108.2 a
Nutrient availability Peri-urban 2,298.4 a 1,819.9 a 478.6 b 100.1 a 705.7 b 21.3 a 5,779.6 b 0.73 b 0.50 b 10.2 a 79.9 b Rural 2,204.5 b 1,811.7 a 392.8 c 85.9 b 645.6 c 18.3 c 5,916.3 b 0.70 c 0.48 c 10.2 a 76.3 b
Population > 20% deficit Urban Peri-urban Rural 15.7 - - 3.5 62.1 11.0 24.5 73.4 91.1 70.2 12.6 13.8 - - 5.1 65.2 13.2 34.9 81.2 96.4 76.5 27.8 20.7 11.2 73.3 18.3 31.7 86.1 96.7 78.7 30.9
The recommended level of nutrient consumption was taken from Food and Nutrition Board (1989), and the deficiencies
4.9.2 Food diversity The diversity in food is universally recognized as a key component of a healthy diet (Hoddinott and Yohanne, 2002). In Hanoi, the consumption of cereals decreased and the consumption of all other foods increased at higher levels of food diversity (Table Table 36. Relationship of food diversity and food and nutrient consumption
Food/nutrient Food consumption Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total food Nutrient Calories Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C Units g g g g g g g g kcal g mg mg IU mg mg mg mg Low 432.7 239.5 35.6 82.0 32.1 10.2 94.1 926.3 2,086.6 78.7 567.8 16.6 5,341.0 0.7 0.4 9.8 73.6 Diversity level Medium 409.8 251.8 71.4 91.9 46.7 22.6 132.4 1,026.5 2,241.2 96.3 704.1 20.1 6,283.5 0.7 0.5 10.1 83.6 High 392.9 268.9 163.1 102.5 59.6 33.4 162.4 1,182.9 2,350.5 108.2 842.5 22.7 7,468.7 0.8 0.6 10.9 104.9
Results and Discussion
36). This increased the consumption of micronutrients, although the consumption of carbohydrates and protein were also enhanced to a certain extent. Food diversity increased from rural to urban areas. The diversity also increased with improved income. The farmer group had lower food diversity than the non-farmer group, as the sample of non-farmer predominantly consisted of urban residents who have greater access to diversified food from the market. The vegetable-farmer households had higher food diversity than non-vegetable farmers, although the difference was not significant (Table 37). Table 37. Food diversity by location, and respondent group
Location/group Location Urban Peri-urban Rural Income group Low Middle Upper Farming-based group Farmer - Vegetable farmer - Non-vegetable farmer Non-farmer Diversity index 5.86 a 4.69 b 4.15 c 4.45 b 5.09 a 5.26 a 4.22 b 4.29 a 4.19 a 5.64 a
4.9.3 Processed food Processed food has value-added properties. Such properties generally make the food more convenient for consumers, but not necessarily more nutritious. Overall. Overall about 8% of food was bought as readymade and 7% of food (mainly fruits) was consumed as fresh. The remaining food passes through the cooking process in the house. The highest proportion of readymade food is in the category of “others”. About 12% of egg and milk and 7% of meat were consumed as readymade in Hanoi. Surprisingly, about 5% of rice was also purchased readymade (Table 38). Region. The proportion of readymade or processed food increased as one moved from rural to urban areas (Table 39). Surprisingly, the opposite was true for fruits. Less than 1% vegetables were demanded as processed on all the three sites. This suggests that urbanization will decrease the demand for processed fruits and increase the share of fresh fruits consumed, and vegetables will continuously be demanded as fresh and then cooked at home. The urbanization will dramatically increase the demand for readymade (or processed) cereals and egg and milk, while having only slight impact on the readymade aquatic products.
An Analysis of Food Demand Patterns in Hanoi
Table 38. Processing stage of food (% of total food) by food group and region
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Home cooked 95.1 98.2 32.3 93.1 99.0 88.2 74.9 84.8 Percentage of total food Ready-made 4.9 0.4 1.6 6.9 1.0 11.8 25.0 8.2 Taken fresh 0.0 1.4 66.1 0.0 0.0 0.0 0.1 7.0
Table 39. Processing stage of food (% of total food) by food group and region
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total HC 87.5 96.5 7.7 90.7 98.6 67.6 70.8 75.7 Urban RM 12.5 0.9 0.8 9.3 1.4 32.4 29.1 9.5 TF 0.0 2.7 91.5 0.0 0.0 0.0 0.1 14.9 Peri-urban HC RM TF 95.2 98.8 13.4 91.4 99.8 87.1 71.2 87.0 4.8 0.3 1.9 8.6 0.2 12.9 28.4 6.9 0.0 0.9 84.7 0.0 0.0 0.0 0.3 6.0 HC 97.8 98.7 25.9 92.5 99.9 97.4 69.0 91.1 Rural RM 2.2 0.5 4.7 7.5 0.1 2.6 30.9 5.7 TF 0.0 0.8 69.4 0.0 0.0 0.0 0.1 3.2
Note: HC = home cooked; RM = readymade or processed; TF = taken fresh.
Income group. As expected, the shares of readymade (or processed) and fresh foods increased as we moved from low to upper income groups, while the opposite was true for the share of food cooked at home, although this difference was relatively small (Table 40). This may partly reflect most women’s desire to spend less time cooking, and partly the changing demand for the processed and fresh uncooked food with increasing income. With varying degree, this is true for almost all individual food groups. Professional group. Farmers consumed a noticeably higher share of food that was cooked at home and a smaller share of readymade and fresh food compared to their counterpart non-farmer group (Table 41). The main difference came in cereals, fruits, and egg and milk where farmers consumed a significantly higher proportion of home-cooked food. In cereals and egg and milk, they consumed a lower proportion of readymade food, while surprisingly they consumed a significantly higher proportion of readymade (or processed) fruits compared to the non-farm group.
Results and Discussion
Table 40. Processing stage of food (% of total) by food group and income level
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total Low income HC RM TF 95.8 98.7 14.9 92.0 99.6 86.5 73.1 88.3 4.2 0.3 1.7 8.0 0.4 13.5 26.6 6.2 0.0 1.0 83.4 0.0 0.0 0.0 0.3 5.5 Middle income HC RM TF 93.1 98.0 11.5 91.1 99.5 79.8 69.7 83.5 6.9 0.6 1.4 8.9 0.5 20.2 30.2 7.9 0.0 1.4 87.1 0.0 0.0 0.0 0.1 8.6 Upper income HC RM TF 92.3 96.6 10.6 91.4 98.8 73.0 65.0 80.0 7.7 1.1 2.4 8.6 1.2 27.0 34.9 8.5 0.0 2.3 87.0 0.0 0.0 0.0 0.1 11.9
Note: HC = home cooked; RM = readymade or processed; TF = taken fresh.
Table 41. Processing stage of food (% of total food) consumed by food group and professional group
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Total HC 97.3 98.9 21.5 92.1 100.0 94.5 70.7 90.5 Farmer RM 2.7 0.4 4.5 7.9 0.0 5.5 29.2 6.0 TF 0.0 0.7 74.0 0.0 0.0 0.0 0.1 3.5 HC 89.4 97.0 8.7 90.9 98.9 71.4 69.8 78.0 Non-farmer RM 10.6 0.7 0.7 9.1 1.1 28.6 29.9 8.9 TF 0.0 2.3 90.6 0.0 0.0 0.0 0.3 13.1
HC = home cooked; RM = readymade or processed; TF = taken fresh.
4.9.4. Food Eaten Outside We estimated that about 9% of the food was eaten outside the house during the survey year for the whole sample. The proportion of food eaten outside was directly related with income; the estimated percentage of food eaten outside was highest in urban areas and lowest in rural areas; vegetable farmers consumed a lower proportion of food outside the house, but the difference was not significant; among the non-farmer groups, the group in the urban area consumed the highest proportion of food outside while the group in the rural area ate the lowest proportion (Table 42). Therefore, with increasing income and urbanization the demand for restaurant food is expected to increase. This will increase the demand for high quality and safe food as people who eat outside care more about these food attributes.
An Analysis of Food Demand Patterns in Hanoi
Table 42. Food eaten outside (%) by region and income and farmer group
Group/locality Income group Low Middle Upper Locality group Urban Peri-urban Rural Professional group Farmer type - Vegetable farmer - Non-vegetable farmer Non-farmer type - Non-farmer in urban - Non-farmer in peri-urban - Non-farmer in rural Overall Food taken outside (%) 7.7 b 9.5 a 10.1 a 13.4 a 9.0 b 5.0 c 5.8 a 6.0 a 13.6 a 11.3 a 4.7 b 8.9
4.9.5 Difference in prices Difference in food prices is a composite measure of food quality which includes the perceived difference of consumers in terms of nutrient, taste, hygienic and safety conditions, convenience in purchase and preparation, etc. Some of the price difference of a given food across regions may also be due to the different marketing cost, especially of transportation and retailing costs, but such differences will be minimal if the regions are closer to each other, such as urban, peri-urban, and rural regions of Hanoi prescribed in this study. The prices differences across regions and income groups were statistically tested for each food item where number of price observations was 10 or more in each group, and results are reported in Appendix 5. The prices of a large number of food commodities increased as we move from rural to urban areas reflecting an improvement in consumers’ perceived quality of these items. For example, out of 71 commodities where a price difference was statistical tested, 54 had significantly higher prices in urban as compared to rural areas, and a similar number of food items had significantly higher prices in urban as compared to in peri-urban areas. A large number of commodities had significantly higher prices in peri-urban compared to rural areas as well. A greater number of fruit and vegetable commodities showed higher prices near the urban center compared to commodities of other food items (Table 43). To quantify the extent of price difference across regions and income groups, we ran a regression of logarithm of price for each food item (where the number of observations was 50 or more) on region, income, and season dummies, and results for price differences across regions and income groups are reported in Table 44. Again the price differences
Results and Discussion
Table 43. Results of the analysis of the price difference across region and income group (number of commodities)
Group Cereals Vegs. and fruits Meats Aquatic prod. Others Total Eq Hi Lo Eq Hi Lo Eq Hi Lo Eq Hi Lo Eq Hi Lo Eq Hi Lo 8 5 6 3 1 3 0 0 0 1 1 0 10 16 14 24 33 28 26 24 22 11 3 16 0 4 1 6 1 5 1 6 1 10 0 8 5 7 2 7 3 6 0 0 2 0 0 0 0 1 0 2 3 4 3 2 2 2 1 1 0 1 12 0 3 11 0 7 8 0 7 7 0 11 5 0 10 7 2 0 1 0 0 0 15 30 28 44 64 56 54 49 40 30 13 33 2 1 4 2 2 0
Region Urban–rural 0 Urban–peri-urban 4 Peri-urban–rural 2 Income group Upper–low 5 Upper–medium 7 Medium–low 6
between urban to rural, urban to peri-urban, and peri-urban to rural groups were positive and significant for 86%, 86% and 76% cases, respectively. The positive price difference across urban and rural areas was highest compared Table 44. Regression analysis of food quality as represented by price difference (%) across region and income group
Commodity Rice Shrimp instant noodle Pork Fresh fish Banana Lemon Kangkong Spring onion Common cabbage Tomato Kohlrabi Hen egg Duck egg Green tea Pickles Iodized salt Cooking oil Pig fat Soybean cake Fish sauce Other alcohols Total positive significant (%) Urban- rural 38.5* 18.7* 30.0* 59.3* 19.8* 26.6* 61.1* 18.7* 40.7* 21.5* 48.3* –1.9 6.1* 23.1* 35.0* –2.9* 4.9* –1.8 14.1* 92.4* 34.1* 85.7 Region Urban- Peri-urban- peri-urban rural 23.6* 6.1* 13.0* 25.9* 1.3 10.5* 32.6* 18.5* 18.5* 15.7* 35.2* 6.6* 5.4* 8.7* 17.3* –1.8* 0.4 2.9* 8.0* 60.4* 16.4* 85.7 14.8* 12.6* 17.0* 33.5* 18.5* 16.2* 28.5* 0.2 22.2* 5.8* 13.1* –8.5* 0.6 14.4* 17.7* –1.1 4.5* –4.7* 6.1* 32.1* 17.7* 76.2 Upper- low 9.7* 8.5* 9.0* 23.3* 5.2 0.5 15.0* 0.9 12.0* 3.4 10.7* 7.6* 1.4 8.9* 19.1* 3.1* 4.4* 0.7 –0.3 31.5* 19.1* 66.7 Income Upper- medium 4.8* 2.1 3.5* 7.2* 3.8 –3.4 4.8* 1.8 0.4 –1.4 0.7 4.7 –0.8 1.9 12.5* 2.4* 1.6* –0.4 0.2 11.2* 14.2* 42.9 Mediumlow 4.9* 6.4* 5.5* 16.2* 1.3 4.0 10.2* –0.9 11.6* 4.8* 10.0* 3.0 2.2 6.9* 6.7* 0.6 2.8* 1.1 -0.5 20.3* 5.0* 61.9
An Analysis of Food Demand Patterns in Hanoi
to the other regions. It ranged from 4.9% for cooking oil to 92.4% for fish sauce. The difference between urban and peri-urban regions ranged between 2.9% in pig fat and 60.4% for fish sauce. Similarly, the difference across peri-urban and rural areas ranged from 4.5% for cooking oil to 33.5% for fresh fish. Such a large difference in food prices across regions can be attributed to only quality in these food items. The differences in prices between upper-low, upper-medium and medium-low income groups were positive and significant for 67%, 43% and 62% cases, respectively. This suggests that as people earn more income they are willing to pay higher prices and purchase better quality food.
4.10 Demand elasticities We first used the Almost Ideal Demand (AID) System at each stage. However, the results of the AID model at the first stage were not consistent. 5 Therefore, we decided to use the Linear Expenditure System (LES) at the food group level, while the AID system was specified for vegetable subgroups.
4.10.1 Elasticities by food group (first stage)
The LES of equation (2) in Appendix 3 was estimated for seven food groups. Most of the estimated coefficients were highly significant statistically. The demand elasticities were estimated using the parameters of equation (2) into equations (4) to (6) of Appendix 3 at each household level and then the average was estimated for different groups, and for the overall sample. The results for the overall sample are reported in Table 45.
Table 45. Demand elasticities of major food groups in Hanoi
P ercentage change in the quantity consumed with one percentage change in the price of: Aquatic Egg and Food group Cereals Vegetables Fruits Meats products milk Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others –0.1361 –0.0068 –0.0036 –0.0007 –0.0029 –0.0018 –0.0026 –0.0145 –0.2158 –0.0025 –0.0005 –0.0020 –0.0012 –0.0018 –0.2513 –0.0817 –0.5467 –0.0088 –0.0345 –0.0214 –0.0312 –0.3819 –0.1242 –0.0669 –0.9892 –0.0524 –0.0326 –0.0475 –0.3256 –0.1058 –0.0570 –0.0114 –0.7017 –0.0278 –0.0405 –0.2695 –0.0876 –0.0472 –0.0095 –0.0370 –0.7053 –0.0335 Others –0.2486 –0.0808 –0.0436 –0.0087 –0.0341 –0.0212 –0.7872 Income 0.0845 0.0629 1.4384 1.7756 1.4739 1.2955 1.0579
Note: The bold numbers show the own price elasticity. We got high cross-price elasticities, mostly positive and near or higher than one, suggesting strong substitutability among food groups, a violation of the separability assumption. It looks that the scenario of allocating at least a subsistence level of expenditure to all food groups and then distributing the remaining expenditure linearly under LES model looks more reasonable for Vietnamese consumers.
5
Results and Discussion
All own-price elasticities had correct (negative) signs. The own-price elasticities of cereals and vegetables are lowest, suggesting that they as each group are an integral part of the diet in Hanoi. This is followed by fruits, which have own-price elasticity at –0.55. Meats have the highest own-price elasticity at around -1.0, suggesting that a 10% raise in its price will bring almost equal reduction in its consumption. Aquatic products and egg and milk foods have similar own-price elasticities at around –0.7. A positive cross-price elasticity suggests that the two commodities are gross substitutes while negative cross-price elasticity implies that they are complementary to one another. The cross price elasticities were generally small (and mostly negative), suggesting separability among food groups (Table 45). All income elasticities have a correct (positive) sign. The income elasticity of vegetables was lower than all other food groups. A 10% raise in income will increase 0.6% of vegetable consumption and 0.8% of cereal, while such increase in other food items will enhance more than 10% of their consumption (Table 45). We estimated the complete set of demand elasticity for three locations (urban, periurban, and rural), and two income groups (low and high6) by using the group-specific parameter and mean value of price and share for each group.7 The own price elasticities, where major differences across these groups were observed, are reported in Table 46 while the full set of the elasticity is reported in Appendix 6. In general, own price elasticities of major food groups decrease as we move from rural to urban areas. In case of aquatic products, however, the peri-urban and rural consumers have similar elasticities. Surprisingly, the income elasticity of fruits decreases as we move from rural to urban areas, while nonsignificant differences in income elasticities of cereals, vegetables and meats across the three regions were observed. The income elasticities of aquatic products were significantly higher in peri-urban areas compared to urban areas and rural areas, while the latter two regions have similar income elasticities. On the other hand, the income elasticity of egg and milk was highest but similar in peri-urban and rural areas (Appendix 6). The high income has significantly lower own-price elasticities for fruits, meats, aquatic products, egg and milk, and others, while the difference was small in case of cereals and vegetables (Table 46). The income elasticities of fruits, aquatic products, and egg and milk were lower for the high income group, while cereals, vegetables, meats, and others had insignificant difference in income elasticities across the income groups.
Initially we estimated the elasticity for three income groups (low, medium, and upper), but no statistical difference was found between medium and high groups, therefore, these two groups were merged. 7 We first estimated the LES model separately for each group, but due to small number of observations in certain groups got biased results for certain parameters. However, where number of observations was not a problem, the elasticities through dummy variable approach and separately estimating the model for each group were very similar.
6
0
An Analysis of Food Demand Patterns in Hanoi
Table 46. Own-price demand elasticities of major food groups by location and income group in Hanoi
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Urban –0.0894 –0.1490 –0.2525 –0.7571 –0.4100 –0.3683 –0.5336 Location Peri-urban –0.1389 –0.2008 –0.6830 –1.0779 –0.9558 –0.9561 –0.8493 Rural –0.2070 –0.2626 –1.0253 –1.1931 –0.8000 –1.1881 –0.9471 Income group Low High –0.1749 –0.2190 –0.7148 –1.1428 –0.9283 –1.0600 –0.8790 –0.1284 –0.2132 –0.5169 –0.9550 –0.6495 –0.6291 –0.7578
4.10.2 Elasticities by vegetable subgroup (second stage)
At the second stage, the demands for various vegetable groups were estimated using the Almost Ideal Demand System specified in equation (7) of Appendix 3. The unconditional elasticities were estimated using equations (16) and (17), and results are reported in Table 47. No statistical differences in own and cross price elasticities were observed across the three regions and two income groups at this stage, therefore, we do not report the elasticities for these regions and income groups. Overall for the whole sample, all own-price and income elasticities were carrying expected signs. Income elasticities for individual vegetables ranged from 0.047 for kangkong to 0.076 for ‘root and stem’ vegetables and pulses. Unlike own-price elasticities for overall vegetable group, the elasticities for individual vegetables were high: higher than –0.45. This is because of the availability of some good substitutes for each individual vegetable group. For example, when price of cucurbits is increased by 10%, its consumption is reduced by 5.7%, while the consumption of leafy brassica is increased by 1.8%, other fruit and flower vegetables by 4.4%, and other leafy vegetables by 0.67%. Similarly cucurbits, other leafy vegetables, and root and stem vegetables may substitute for other fruit and flower vegetables. Other complementary (negative sign other than own-price elasticity) and substitute (positive sign) relationships between different vegetable groups can be seen in Table 47. Out of 42 cross-price elasticities, 23 were carrying positive signs, implying that substitute relationships dominated among vegetable subgroups although its extent varied by vegetable type.
Results and Discussion
Table 47. Demand elasticities by vegetable group
Vegetable Kangkong Root and stem Leafy brassica Cucurbits Other fruits and flower Other leafy Pulses Percentage change in the quantity consumed with one percentage change in the price of: Root and Leafy Other fruits Other Kangkong stem brassica Cucurbits and flower leafy Pulses Income –0.4654 –0.0265 0.1227 –0.0774 –0.0096 –0.1218 –0.1930 –0.1549 –0.8135 –0.0224 –0.0521 0.0644 0.0943 0.2565 0.2782 0.0055 –0.5351 0.0877 –0.0661 0.0628 0.1122 –0.0073 –0.0179 0.1849 –0.5733 0.4431 0.0672 –0.1549 –0.1779 0.2086 –0.1735 0.4298 –1.1103 0.3862 –0.0359 –0.0582 0.1861 0.0979 0.0452 0.2693 –0.8490 0.2652 –0.2283 0.5212 0.2296 –0.1212 –0.0150 0.3373 –0.6606 0.0470 0.0760 0.0614 0.0605 0.0625 0.0686 0.0759
Note: The bold numbers are own-price elasticities.
An Analysis of Food Demand Patterns in Hanoi
5 Summary and Policy Implications
This study provides a comprehensive quantitative sketch of the consumption pattern of food in Hanoi and its surrounding rural areas. A comprehensive collection of household data through a 24-hr recall method in a repeated survey over three seasons enabled us to quantitatively compare the food and nutrient consumption patterns in urban, peri-urban, and rural areas of Hanoi across various socioeconomic groups and seasons. The quality of food demanded by consumers was also analyzed in terms of its nutrient supply, diversity, extent of food eaten outside, stage of its processing, and prices. The relative contribution of different food sources were quantified highlighting the importance of farm and home garden production as well as various other food market types. At the end we estimated the price and income elasticities of various food and vegetable groups as well as for major nutrients important for human health. It is expected that this quantitative analysis of food and its quality, and demand elasticities will help policy makers predicting the food demand changes in Hanoi, both in quality and quantity, and formulate efficient food policies in the wake of various shocks in the domestic and international markets. Currently cereals dominate in the dietary pattern of Hanoi, at least in terms of food quantity, and vegetables occupy the second place in this pattern. Vegetables are an important source of micronutrients especially for low-income households. The demand of vegetables will also increase with higher incomes and urbanization. But the rate of increase in vegetable demand would be far less than the increase in livestock products. The high increase in livestock products will increase the contribution of a fat-based source for calories. This should be a cause of concern, as it may lead to obesity. Moreover, the larger concentrations of animals in peri-urban areas to meet these demands have led to pollution problems (Delgado et al. 1999). This creates an opportunity for public policies to appropriately manage not only the livestock demand but also its production spaces in urban areas to maintain public health and the environment. Farms are the major source of food supplies in rural areas of Hanoi and second major source for peri-urban areas; low-income households of the city rely heavily on this source. Therefore, strengthening agricultural production in and around Hanoi city through technological innovation can contribute greatly in enhancing food security, especially for low-income households in Hanoi. The contribution of home gardens in rural areas is even higher than that of farm production in supplying vegetables, fruits, meats, and egg and milk. The low-income group in the city relies more heavily on home garden sources. As urbanization decreases the home garden share in food supply, it will have consequences on the fresh food supply in urban areas in general, and for food security among low-income families. Therefore, promotion of home gardening in rural and peri-urban areas can improve food security and fresh food supplies especially among the poor farm families of Hanoi. The temporary informal markets turned out to be the major source of food supplies in urban areas. Policy makers normally consider these markets as problem for the city
Summary and Policy Implications
and a source of contaminated food. Needless restrictions are often imposed on the sector, which simply increase the cost of food. Since this is an important source of not only food but also employment, the government should integrate this sector into formal markets by providing appropriate space and licensing and equipping them with appropriate tools and skills for keeping the food hygienic. Enhancing the food processing skills of the people engaged in this informal food marketing can also help them to integrate into the formal food market. Contrary to common perception, overall food consumption was highest during the hot-wet season and lowest during the cold-wet season. The more food consumption during the hot-wet season, which is against the restricted food supply notion, may be due to an increased requirement of the human body during this season. The consumption of vegetables was lowest and their prices highest during the hot-wet season (as they are difficult to produce during this season), but their low level of consumption was compensated by the high consumption of tropical fruits. Moreover, the consumption of leafy vegetables, which can favorably be grown during the hot-wet season, was also enhanced during this season. Therefore, seasonality in food supply did not have much impact in terms of nutrient availability. The implication is that policy makers should not be concerned on reducing seasonality of particular components of Hanoi’s food supply; rather they should focus on improving the overall food supply to the city wherever and whenever it is possible. Trade and efficient production systems are the tools to tackle seasonality in food supply. Agriculturalists should focus to resolve the production problems of hotwet season crops, such as leafy vegetables and tropical fruits, rather than overcoming seasonality in a particular food item. The trade-oriented policies should try to link the efficient production zones in hot-wet season with demand centers. About 9% of the food was consumed outside the home, and the proportion was higher in urban areas in the upper-income group. Another 7% was bought as cooked or readymade food, again with the proportion higher for urban areas and the upper income group. Therefore, urbanization and increased income are expected to change the food market systems in Hanoi. More demand for restaurant and readymade food is expected in the near future. This may have implications for women labor engaged in household cooking. Moreover, health and food safety issues related to food consumption will surface more strongly. The planners have to adjust the policy environment to avoid the negative implications of these changes. On average, there seems to be no serious nutrient deficiency in Hanoi and its surrounding population. But looking at individual families on a daily basis, a large number of families fall below the daily recommended levels of calcium and vitamins B1, B2, and niacin. A small proportion of population is also deficient in calories and vitamins A and C. All the regions and income levels have these deficiencies, although to a small extent it varies across income groups and regions. This suggests that lack of income is not the only cause of nutrient deficiency. Lack of consumer awareness is also part of the problem. Therefore, efforts are required to improve consumer awareness of healthy diets. Policies can also play an important role in alleviating micronutrient deficiencies.
An Analysis of Food Demand Patterns in Hanoi
For example, reducing the prices of seafood can reduce calcium deficiency, and reducing the prices of vegetables, especially leafy, fruit and root vegetables, will mitigate the vitamin A deficiency significantly. Our analysis suggests that urbanization and enhanced incomes in urban areas will bring structural and qualitative changes in food consumption and distribution patterns. Consumers will demand more quality food in terms of its micronutrient density and diversity, and will be willing to pay higher prices for food quality. The role of livestock products, fruits, and aquatic food is expected to increase. The demand for readymade and restaurant food is expected to increase in the near future. The fresh food supply from home gardens and peri-urban farms may diminish, and supermarkets may become more important than street vendors. These changes will have different implications for different sectors of the society. It is expected that the analysis conducted in this study will help policy makers in preparing for and mitigating the negative impacts of these changes in food markets on health, food security of the poor, and hygienic conditions of food distribution.
References
References
Adrian, J. and R. Daniel. 1976. Impact of socioeconomic factors on consumption of selected food nutrients in the United States. Amer. J. Agr. Econ. 58:31–38. Ali, M., S.N. Wu, and M.H. Wu. 2000. Estimation of demand elasticities for main food items in Taiwan. Asian Vegetable Research and Development Center, Shanhua, Taiwan. Ali, M. and U. Farooq. 2004. Dietary diversity to enhance rural labor productivity: evidence from Pakistan. Paper presented in the Annual Meeting of American Agricultural Economic Association (AAEA), 1–4 July 2004, Denver, Colorado, USA. Basiotis, P., M. Brown, S.R. Johnson, and K.J. Morgan. 1983. Nutrient availability, food costs and food stamps. Amer. J. Agr. Econ. 65:685–693. Blanciforti, L.A., R.D. Green, and G.A. King. 1986. US consumer behavior over the postwar period: an almost ideal demand system analysis. University of California, Davis, Giannini Foundation Monograph No. 40, August 1986. Bridier, B.E. 2000. Etude de la perception de la qualité des légumes par les consommateurs de Hanoi (Vietnam). Cas particulier de la qualité sanitaire. ESAT “Développement agricole tropical. Option: Valorisation des productions”. Montpellier, CNEARC: 149. Bussink, W.C.F. 1970. A complete set of consumption coefficients for West Pakistan. The Pakistan Development Review 10(2): 193–231. Cheema, A.A. and M.H. Malik. 1985. Changes in consumption patterns and employment under alternative income distribution in Pakistan. The Pakistan Development Review 24(4): 1–22. Chen, K. and C. Chen. 2000. Cross product censoring in a demand system with limited dependent variables: a multivariate probit model approach. Staff Paper No. 00-02, Department of Rural Economy, Faculty of Agriculture, Forestry and Home Economics, University of Alberta, Edmonton, Canada. Deaton, A. and J. Muellbauer. 1980a. An almost ideal demand system. American Economic Review 70(3):312–326. Deaton, A. and J. Muellbauer. 1980b. Economics and consumer behavior. Cambridge University Press, p. 123–124. Delgado, C., M. Rosegrant, H. Steinfeld, S. Ehui, and C. Courbois. 1999. Livestock to 2020: the next food revolution. (Brief) International Food Policy Research Institute, Washington, D.C. Devaney, B. and T. Fraker. 1989. The dietary impacts of the school breakfast program. Amer. J. Agr. Econ. 71:932–948.
An Analysis of Food Demand Patterns in Hanoi
Escalante, C. and P. Barry. 2001. Farm-level evidence on the risk balancing hypothesis from Illinois grain farms. Paper presented at 2001 American Agricultural Economics Association Annual Meeting, 5–8 August 2001, Chicago, Illinois. Fan, S., E.J. Wailes, and G.L. Cramer. 1995. Household demand in rural China: A twostage LES-AIDS model. Amer. J. Agr. Econ. 77(1):54–62. FAO (Food and Agriculture Organization). 1996. Synthesis of the technical background documents for the World Food Summit. 13–17 November 1996, FAO, Rome, Italy. Farooq, U. and M. Ali. 2003. Combating micronutrient deficiency in Pakistan by increased vegetable use. AVRDC – The World Vegetable Center, Shanhua, Taiwan. Farooq, U. and M. Ali. 2005. Combating micronutrient deficiency by increased vegetable use in Pakistan: Will the poor benefit also? AVRDC – The World Vegetable Center, Shanhua, Taiwan. Figuié, M. 2003. Vegetable consumption behaviour in Vietnam. Sustainable Development of Peri-urban Agriculture in South-east Asia (Susper) Project, Hanoi http://www.cipotato. org/urbanharvest/documents/CIRAD_Susper.pdf downloaded on 13 July 2005. FIRDI and PUST (Food Industry Research and Development Institute and Pintung University of Science and Technology). 1998. Food composition table in Taiwan area. Department of Health, Executive Yuan, Taipei, Taiwan. Food and Nutrition Board. 1989. Recommended dietary allowances. 10th ed. Washington D.C.: National Academy Press. Gao, X.M., E.J. Wailes, and G.L. Cramer. 1995. A microeconometric model analysis of U.S. consumer demand for alcoholic beverages. Applied Economics 27(1):59–69. Gao, X.M., E.J. Wailes, and G.L. Cramer. 1996. A two-stage rural household demand analysis: microdata evidence from Jiangsu Province, China. Amer. J. Agr. Econ. 78(3):604–613. Gould, B.W., T.L. Cox, and F. Perali. 1991. Demand for food, fats and oils: the role of demographic variables and government donations. Amer. J. Agr. Econ. 73:212–221. Haden, K. 1990. The demand for cigarettes in Japan. Amer. J. Agr. Econ. 72(2):446– 450. Han, T. and T.I. Wahl. 1998. China’s rural household demand for fruit and vegetables. Journal of Agriculture and Applied Economics 30(1):141–150. Hannah, L. and J.A. Kay. 1977. Concentration in modern industry: theory, measurement, and UK experience. London: Macmillan Press. Hanson, P. and P. Simons. 1995. Measures of buyer concentration in the Australian wool market. Rev. Mktg. and Agr. Econ. 63:304–310. Heien, D. and G. Pompelli. 1988. The demand for beef products: cross-section estima-
References
tion of demographic and economic effects. Western Journal of Agricultural Economics 13(1): 37–44. Heien, D. and C.R. Wessells. 1990. Demand systems estimation with microdata: A censord regression approach. Journal of Business and Economic Statistics 8(3):365–371. Hoddinott, J. and Y. Yohanne. 2002. Is dietary diversity an indicator of food security or dietary quality? A review of measurement issues and research needs. FCND Discussion Paper Brief No. 140. IFPRI, Washington DC. Liu, K.E. and W.S. Chern. 2001. Effects of model specification and demographic variables on food consumption: Micro data evidence from Jiangsu, China. Paper presented at 11th Annual World Food and Agribusiness Forum, World Food and Agribusiness Symposium of the International Food and Agribusiness Management Association, 27–28 June 2001, Sydney, Australia. Merrill, A.L. and B.K. Watt. 1995. Application of general factors to national food supplies. In: Energy value of foods, USDA, Agriculture Handbook 74:48–49. Pitt, M.M. 1983. Food preferences and nutrition in rural Bangladesh. Rev. Econ. and Statist. 65:105–114. Pollak, R.A. and T.J. Wales. 1978. Estimation of complete demand systems from household budget data: The linear and quadratic expenditure systems. American Economic Review 68:348–359. Sadoulet, E. and A. de-Janvry. 1995. Quantitative development policy analysis. Baltimore and London: The John Hopkins University Press. Sahn, D.E. 1988. The effect of price and income changes on food-energy intake in Sri Lanka. Econ. Develop. and Cultur. Change 63:315–340. Shonkwiler, J.S. and S.T. Yen. 1999. Two-step estimation of a censored system of equations. Amer. J. Agr. Econ. 81(4):972–982. Tan Loc, N.T. 2002. Le développement des magasins et des supermarchés dans la filière des légumes frais à Hanoi et Ho Chi Min, RIFAV/ CIRAD: 21. Tauer, L.W. and T.B. Seleka. 1994. Agricultural diversity and cash receipt variability for individual states. Cornell Agricultural Economics Staff Paper No. 94-1, Department of Agricultural, Resource and Managerial Economics, College of Agriculture and Life Sciences, Ithaca New York: Cornell University. Thuy, N.T.T., M.H. Wu, and T.V. Lai. 2002. North Vietnam. In: Ali, M. (ed.). The vegetable sector in Indochina countries: farm and household perspectives on poverty alleviation. Technical Bulletin No. 27, Bangkok, Thailand: Asian Vegetable Research and Development Center-Asian Regional Center. Vijayalakshmi, P., S. Amirthaveni, R.P. Devadas, K. Weinberger, S.C.S. Tsou, and S. Shanmugasundaram. 2003. Enhanced bioavailability of iron from mungbeans and its
An Analysis of Food Demand Patterns in Hanoi
effects on health of schoolchildren. Technical Bulletin No. 30, Publication Number 03559. AVRDC—the World Vegetable Center, Shanhua, Taiwan. 32 pp. Wang, J., X.M. Gao, E.J. Wailes, and G.L. Cramer. 1996. U.S. consumer demand for alcoholic beverages: cross-section estimation of demographic and economic effects. Review of Agricultural Economics 18(3):477–489. Weinberger, K.M. 2001. Combating micronutrient deficiencies of the Indian poor: do technological innovations in vegetables and pulses research present a viable choice? AVRDC–The World Vegetable Center, Shanhua, Taiwan. Wu, Y.R., E. Li, and S.N. Samuel. 1995. Food consumption in urban China: an empirical analysis. Applied Economics 27(6):509–515.
Appendixes
Appendix 1
The list of sample provinces, districts and communes/wards
Area Urban Hanoi District Cau Giay Hai Ba Trung Tay Ho Hoan Kiem Thanh Xuan Dong Da Ba Dinh Peri-urban Hanoi Tu Liem Thanh Tri Gia Lam Soc Son Dong Anh Hung-Yen (rural) Kim Dong Tien Lu An Thi Ha Tay (rural) Hoai Duc Phu Xuyen My Duc Commune/ward Trung Hoa, Quan Hoa Mai Dong, Truong Dinh, Pham Dinh Ho, Bach Mai (46), Thanh Nhan, Dong Mac Thuy Khue (33), Xuan La Dong Xuan, Tran Hung Dao, Phuc Tan Nhan Chinh, Kim Giang, Khuong Mai Cat Linh, Quang Trung, Quoc Tu Giam, Khuong Thuong, Trung Phung, Lang Ha Ngoc Ha, Nguyen Trung Truc, Cong Vi Lien Mac, Trung Van, Tay Mo, Tay Tuu, Xuan Phuong Linh Lam, Dinh Cong, Thinh Liet, Ta Thanh Oai Cu Khoi, Kieu Ki, Thuong Thanh, Ninh Hiep, Sai Dong, TT Yen Vien, TT Duc Giang Tan Dan, Tan Hung, Mai Dinh, Minh Phu Viet Hung, Van Ha, Nam Hong, Mai Lam, Duc Tu Duc Hop, Luong Bang, Toan Thang, Tho Vinh, Chinh Nghia Duc Thang, Le Xa, Cuong Chinh, Thu Sy, Hung Dao Phu Ung, Dao Duong, Xuan Truc, Van Nhue, Da Loc Van Canh, An Khanh, Duong Lieu, Cat Que, Duong Noi Phu Tuc, Hong Minh, Son Ha, Nam Phong, Dai Xuyen Phuc Lam, An Phu, Le Thanh, Dai Hung, Hung Tien
0
An Analysis of Food Demand Patterns in Hanoi
Appendix 2
Grouping of food commodities
Cereals Vegetables Rice, corn/maize, bread, potato, sweet potato, taro, cassava, glutinous rice, instant noodle, fresh rice noodles, shrimp instant noodle, arrowroot noodles, rice vermicelli, square glutinous rice cake, kind of rice cake, rice flour, steamed rolled rice pancake, kind of fried cake, glutinous rice cake, rice sheet/ rice paper, instant soup, steamed glutinous rice, floating cake Allium Other root and stem Heading cole Fruits Cucurbits Other fruit and flower Leafy Pulses Tropical fruits Subtropical fruits Temperate fruits Meats Aquatic products Egg and milk Others Pork, bones of pig, blood of pig, liver of pig, kidney of pig, insides of a pig, tongue of pig, pudding of pig, feet/leg of pig, pork-pie, meat pie, salted shredded meat, canned meat, meat roll, sausage, Chinese sausage, beef (thigh), beef ( other parts), stomach of cow, cow hooves, chicken, duck, goose, swan, pigeon, other bird, leg/feet of dog, dog meat (other parts), processed meat, pig fat Fresh fish, sea fish, dried/processed fish, fresh shrimps, sea shrimps, dried/ processed shrimps, fresh crabs, fresh shells, snail, fried fish Cow milk, powder milk, other milk, hen egg, duck egg, other egg, butter Black tea, green tea, coffee, sugar, vinegar, pickles, juice, soda (Coca-cola, 7-Up, etc.), iodine salt, glutamate, other salt, cooking oil, sweet cake, tofu, biscuits, fish sauce, black pepper, soyasauce, beers, wines, alcohol, eugenia, galingale, shrimp paste, candy, silk worm cocoon, saffron, chutney, artemisia, absinth, ferment, cat’s ear, pie, milk from soybean, lotus seed, dumpling, pyramidal rice dumpling Spring onion, bulb onion, garlic, leek, celery, spring onion bulb Turnips, carrot, radish, ginger, mushroom, kohlrabi, flower of banana, bamboo shoot Cauliflower, Chinese cabbage, common cabbage Wax gourd, bottle gourd, bitter gourd, cucumber, pumpkin, pumpkin buds, smooth loofah, chayote Sweet pepper, hot chilies, eggplant, tomato, tamarind, baby corn, dracontomelum Kangkong, amaranth, Indian spinach, Chinese mustard, pak-choi, lettuce, star gooseberry, coriander, salad, garland chrysanthemum, leafy vegetable, sweet potato bud, fennel, persicaria, basil, perilla, marjoram, water dropwort, lemongrass, eryngium, pot herbs, Indian taro, cress French bean, mungbean, black bean, peanuts, sesame seed, green peas, yardlong bean Mango, banana, sapodilla, papaya, persimmon, jackfruit, custard apple, star fruit, pineapple, watermelon, muskmelon, pear-shaped melon, casaba melon, coconut, dragonfruit, guava, rambutan, rose-apple, pachyrrhizus, momordica, oleaster, star apple, wampee, sugarcane Orange, longan, pumelo, mandarin, lemon, kumquat
Appendixes
Appendix 3
Two-Stage Demand Estimation
Two approaches are used for estimating demand elasticities. The first approach consists of specifying estimable single-equation demand functions in a pragmatic fashion without recourse to economic theory. This approach was utilized in a number of studies in 1970s and 1980s, for instance, Bussink (1970), Cheema and Malik (1985), and rarely used recently, for instance, Vijayalakshmi et al. (2003). In the second approach, named the complete demand system, all products are treated symmetrically and simultaneously. Using this approach, a number of demand systems have been developed during the 1990s using the primal or dual functional forms. Two of them have become very popular because of their relative expediency. These are the Linear Expenditure System (LES) developed by Stone in 1954 and the Almost Ideal Demand System (AIDS) developed by Deaton and Muellbauer (1980a). We used the complete demand system in our analysis. While estimating a complete system of demand equations, a common problem is that the number of parameters to be estimated rises rapidly with the increase in number of variables in the analysis, thus gradually becoming difficult to estimate. One solution of this problem is to use the two-stage budgeting analysis. Under the two-stage approach, consumers’ budget allocation for food is broken down into two stages. In the first stage, total disposable income is allocated to broad commodity groups, while in the second stage, fixed group expenditures are further allocated over individual goods. Both of these allocations have to be perfect in the sense that the results of the two-stage budgeting must be identical to the results if the allocations were made in one step with complete information. Weak separability of the direct utility function over the broad groups is both necessary and sufficient condition for the second stage of two-stage budgeting (Deaton and Muellbauer, 1980b). The two-stage budgeting analysis is now gradually becoming popular in empirical studies, e.g., Haden (1990), Fan et al. (1995), Wu et al. (1995), Gao et al. (1995 and 1996), Wang et al. (1996), Han and Wahl (1998), Ali et al. (2000), and Weinberger (2001). The Linear Estimation System (LES) is used only for the first stage, while the Almost Ideal Demand System (AIDS) was found employed at both the first and second stages (Table A1). In the present exercise, LES was chosen for the first stage because of the consistency of results obtained using this approach. Following the normal practice in the literature, the AID system was applied at the second stage. Application of LES requires a larger price variation, which can be easily available at the group level (Sadoulet and de-Janvry, 1995), especially in our case where a large data set spread out over three seasons is used. However, LES rules out inferior good and Hicksian complementarities, but these conditions are less likely to be applicable at the food-group level. At the second stage, AIDS was chosen because it permits a full range of commodities, including complementary and substitute goods as well as normal and inferior goods). Moreover, AIDS has theoretical superiority as being flexible in allowing, but not requiring, the general restrictions of demand theory to hold.
An Analysis of Food Demand Patterns in Hanoi
Table A1. Model selection for the first and second stage analysis in two-stage budgeting
Study and year Haden (1990) Fan et al. (1995) Wu et al. (1993) Wu et al. (1995) Gao et al. (1995) Gao et al. (1996) Wang et al. (1996) Han and Wahl (1998) Ali et al. (2000) Weinberger (2001) Country Japan China China China USA China USA China Taiwan India First stage analysis Linear functional form with one year lagged budget shares as habit Linear Expenditure System (LES) AIDS AIDS Gamma-Tobit-Model Generalized Linear Expend- iture System (GLES) Double-hurdle model LES LES LES Second stage analysis Almost Ideal Demand System (AIDS) AIDS AIDS AIDS Combination of Rotterdam, Central Bureau of Statistics (CBS) model and AIDS AIDS
Combination of CBS model and AIDS Linear Approximation of Almost Ideal Demand System (LA/AIDS) AIDS AIDS
Source: author’s compilation.
Linear Expenditure System (LES). For the Ith7 food group, the following functional form of LES was estimated for the first stage allocation analysis,
PI X I = PI gI + b I ( M − ∑ PJ gJ )
J =1
N
(1)
Following Liu and Chern (2001), g I = g I* +
∑d
h =1
s
Ih
H h , where P is the price, X is
the quantity consumed, M is the total per capita food consumption expenditure, N is total number of food groups, Hh pertains to household characteristics8, and g *,β, and δ are parameters to be estimated.
7
In the subsequent text, I, J and i, j refer to food groups and individual commodities/sub-groups within a food group, respectively.
Appendixes
The equation (1) states that the expenditure on the Ith food group is equal to a certain basic (or subsistence) amount of its consumption γ *1 valued at current prices plus a certain proportion bI of total expenditure less the committed subsistence expenditure on all food groups. The consumer first uses up a certain amount of total expenditure in acquiring the subsistence quantities of the food groups γ * =(γ *1, γ *2, ……… γ *n) at current prices and then distribute the remainder over the set of available food groups in a certain fixed proportion explained by the elements of β = (β1, β2,………β n). Estimating demand elasticities from the household cross-section data used in this study presents a major estimation problem because many households consumed zero amounts of various commodities under consideration. Heien and Wessells (1990) proposed a two-step estimation procedure for a demand system with limited dependent variables to resolve this problem. In this procedure a regressor derived from probit estimate in a separate step, called Inverse Mills Ratio (IMR), is augmented in each demand equation. This procedure has increasingly become popular in applied demand analysis (Chen and Chen, 2000). However, Shonkwiler and Yen (1999) pointed out an internal inconsistency in the estimators obtained using Heien and Wessells procedure. Therefore, they proposed a slightly different two-step estimation procedure in which the denominator part (i.e., cumulative-probability function) of Inverse Mills Ratio is multiplied with all exogenous variables and numerator part (i.e. density function) is used as additional variable in the augmented regression as follows:
' PI X I = PI Φ ( z I' s I )g I + b I [ M − ∑ PJ Φ ( z J s J )g J ] + q I f ( z I' s I ) J =1 N
(2)
where zI and sI are the vectors of exogenous variables and parameters respectively in the estimation of Inverse Mills Ratio.9 Symmetry and adding up restrictions are imposed on the system as follows: 0 < bI < 1, åbI = 1, XI > g*1 > 0,
(3)
The adding up property of LES implies that in the above system of equations is one equation redundant, therefore, has to be dropped arbitrarily10. The income, own and cross-price elasticities were computed by using the following equations: Income elasticity:
hI =
bI wI
(4)
Following Heien and Pompelli (1988), the household characteristics were included in the demand equation. This helps to overcome the bias in parameter estimation due to heteroskedosticity and autocorrelation (Fan et al., 1995). The household characteristics included in this analysis are education of the housewife, and dummy variables for season, location and income group. 9 The variables included in the Inverse Mills Ratio were family size, education, dummy variables for the ownership of refrigerator and one-income group. 10 The values of b for the dropped equation were recovered by using equation (2).
8
An Analysis of Food Demand Patterns in Hanoi
Own-price elasticity:
e II = −1 +
(1 − b I ) PI g I PI X I
(5)
Cross-price elasticity:
e IJ =
− b I PJ g J PI X I
(6)
where ω1 is the budget share of Ith food group. ω1, P1X1, and PI are fixed at the meanlevel while estimating elasticities. Regarding the effect of household characteristics on own-price elasticity, if δ coefficient carries a negative sign, it will decrease the value of γ, leading to lowering the estimate of second component of equation (5), which leads to increase the resultant magnitude (in absolute value) of the own-price elasticity. In case of cross-price elasticity, the negative δ coefficient will decrease the value of cross-price elasticity in equation 6. As no γ is involved in income elasticity equation, therefore, the signs of g will not cause any effect. Almost Ideal Demand System (AIDS). The standard AID model was adjusted to overcome the zero-observation problem as suggested by Shonkwiler and Yen (1999) as follows:
M wi , I = a i , I + ∑ gij , I ln p j , I Φ( z i' , I s i ) + b i , I ln I + q i , I f ( z i' , I s i ) PI i , j =1 I
n
s
(7)
where ωi,I represent budget share for the ith subgroup in the Ith food group, n is the total
h =1 number of subgroups in a given food-group, where Hh pertains 11 to the household characteristics ; p is price of sub-food group, except that it is for the ith commodity in the Ith group here; MI is total per capita consumption expenditure; zi,I and si are the vectors of exogenous variables and parameters respectively in the estimation of Inverse Mills Ratio; and PII is the price index for the group under consideration, defined by:
a i , I = a i*, I + ∑ dih , I H h
ln PI I = a o , I + ∑ a i , I ln pi , I + 1 2 ∑
i , I =1
n
n
i , I =1
j , I =1
∑g
n
ij , I
ln pi , I ln p j , I
(8)
The α*, γ, β, δ, and θ are parameters, to be estimated, with the following restrictions:
11
The household characteristics included in this analysis are education of the housewife, dummy variables for season, location and income group.
Appendixes
Adding up:
i , I =1
∑a
n
i, I
=1
i , I =1
∑b
n
i, I
=0
n
∑g
i =1
n
ij , I
=0
∑d
i =1
n
ij , I
= 0 (h = 1 ....s)
(9)
Homogeneity:
j , I =1
∑g
ij , I
=0
(10) (11)
Symmetry:
γij,1 = γji,1
Since the budget shares add up to unity, therefore, during estimation one share equation was arbitrarily dropped to make the system non-singular. The effect of household characteristics was incorporated in the model by allowing the intercept in equation (5) to be a function of these characteristics, therefore:
a = a i*, I + ∑ d ih , I H h
h =1
s
(12)
Following Blanciforti et al. (1986), the uncompensated income, own-price and cross-price elasticities were computed from the parameter estimates using the following expressions: Income elasticity:
h i,I =
b i,I wi,I
+1
(13)
Own-price elasticity:
e ii , I = −1 + g ij , I wi ,I
gij , I wi , I −
−
b i, I a i ,I wi , I −
−
bi,I wi , I
∑
J ,I
gij , I ln p j , I
(14)
Cross-price elasticity: e ij , I =
b i,I a i,I wi ,I
b i ,I wi ,I
∑
J ,I
gij , I ln p j , I
(15)
The unconditional price and income elasticities of individual vegetables in the vegetable group were calculated as: Price elasticity:
* e ij , I = e ij , I + h i , I w j , I (1 + e II )
(16) (17)
Income elasticity:
h i*, I = h i , Ih I
An Analysis of Food Demand Patterns in Hanoi
The equations (7) and (8) along with the restrictions in (9–11) were estimated using Full Information Maximum Likelihood (FIML12) method of SAS version 6.2. In the first iteration of equation (7), the value of P was computed using the Stone price index. In the next iteration, new price index was computed by incorporating the estimated parameter of equation (7) into (8). The AID system was re-estimated using the new price index. The iteration process was repeated until the parameters in (7) were converged.
12
FIML assumes normal distribution of error and generally requires large data. The danger of non-normal distribution of error term may be more common in small data sets. Since our data set is quite large, therefore, the danger of poor results can be ruled out.
Appendixes
Appendix 4
Nutrient source by food and income group
Nutrient Calories Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C Calories Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C Calories Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C Calories Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C Calories Protein Calcium Iron Vitamin A Vitamin B1 Vitamin B2 Niacin Vitamin C Total 2048.8 91.2 566.9 18.3 6,117.9 0.8 0.6 11.8 72.4 2,071.7 100.7 643.3 19.8 7,106.8 0.9 0.8 13.1 81.5 2,115.7 102.2 650.2 20.2 7,164.0 0.9 0.8 13.2 90.9 2,083.8 101.3 652.7 19.5 7,399.5 0.9 0.8 13.4 92.6 2,099 101.8 700.5 19.7 8,292.2 1 0.9 14.8 108.3 Cereals 70.3 38.2 23.2 31.7 0.0 50.0 0.0 58.5 1.4 65.5 34.7 19.8 28.3 0.0 44.4 12.5 51.1 0.7 63.7 34.1 19.3 27.2 0.0 44.4 12.5 49.2 0.7 64.6 34.3 19.4 28.2 0.0 44.4 12.5 49.3 0.5 62.8 34.0 17.7 27.9 0.0 40.0 11.1 43.9 0.5 Percentage supplied by Aquatic Vegetables Fruits Meats products 3.5 6.5 27.0 15.3 87.1 25.0 33.3 14.4 84.9 3.3 6.1 25.0 14.6 79.5 22.2 25.0 12.2 82.1 3.8 6.4 25.5 15.8 77.9 22.2 25.0 13.6 76.9 Low income 0.7 0.3 1.9 2.2 5.0 0.0 0.0 0.8 12.3 Low-mid income 1.3 0.6 2.5 3.0 13.3 0.0 0.0 1.5 15.7 Middle income 1.5 0.7 3.7 4.0 14.1 0.0 0.0 2.3 20.9 4.1 9.9 1.0 7.7 1.8 12.5 50.0 21.2 0.4 6.6 13.5 1.3 11.1 1.7 22.2 50.0 28.2 0.6 7.6 14.1 1.5 11.9 1.4 22.2 50.0 28.8 0.8 6.4 13.8 1.3 11.3 1.8 22.2 50.0 28.4 0.6 9.0 17.6 1.7 14.7 2.0 30.0 55.6 32.4 0.8 1.3 4.6 9.1 0.5 0.0 0.0 0.0 1.7 0.0 1.9 6.5 14.7 1.0 0.0 0.0 0.0 3.1 0.0 1.9 6.4 13.2 1.0 0.0 0.0 0.0 3.0 0.0 2.1 7.3 14.3 1.0 0.0 0.0 0.0 3.0 0.0 2.2 7.4 19.8 1.5 0.0 0.0 0.0 4.1 0.0 Egg and milk 2.4 2.0 2.1 2.2 5.0 0.0 0.0 0.0 0.0 2.8 2.1 2.5 2.0 4.5 0.0 12.5 0.0 0.1 2.8 2.4 3.3 2.5 5.7 0.0 12.5 0.0 0.1 3.5 2.8 3.9 2.6 5.7 0.0 12.5 0.0 0.1 3.4 2.8 3.9 2.5 5.2 0.0 11.1 0.0 0.1 Others 17.9 38.5 35.6 39.9 1.0 0.0 0.0 2.5 0.7 18.5 36.6 34.1 39.4 0.9 0.0 0.0 3.8 0.9 18.7 36.0 33.5 38.1 0.9 0.0 0.0 3.0 0.8 18.5 35.1 32.0 38.5 0.8 0.0 0.0 3.7 0.6 16.5 30.4 25.8 33.0 0.6 0.0 0.0 4.1 0.6
Mid-upper income 3.5 1.4 6.1 0.7 25.3 3.8 15.4 2.6 75.9 15.7 22.2 0.0 25.0 0.0 13.4 2.2 77.5 20.5 4.0 6.8 25.8 16.8 73.8 20.0 22.2 12.8 71.2 Upper income 2.1 1.0 5.3 3.6 18.3 0.0 0.0 2.7 26.8
An Analysis of Food Demand Patterns in Hanoi
Appendix 5
Prices of individual food items (000 VND/kg) by location and income group
Food name Location Urban Peri-urban Rural 3.3 b 3.3 12.1 ab 2.3 b 1.0 1.5 - 6.0 b 25.0 2.8 12.5 b 5.9 a 2.7 a 10.8 15.0 - 28.9 a 5.3 27.3 b 11.2 ab 4.0 b 68.0 ab 12.3 5.3 13.0 33.8 10.7 a 10.0 8.3 - 3.0 b 10.0 10.2 b 8.0 9.9 a 4.0 22.3 b 16.9 b 5.0 12.0 50.0 11.3 a 45.0 - 15.0 56.3 41.7 a 15.0 25.0 23.4 b 14.5 2.9 c 2.2 11.3 b 2.2 b 3.9 2.0 1.0 5.0 12.5 2.8 11.4 c 4.2 b 2.4 b 10.0 - - 25.0 4.2 25.3 10.5 b 3.1c 60.9 b 10.8 6.5 10.0 - 11.0 - 7.8 6.3 3.3 ab 11.0 10.1 b 12.0 8.8 b - 19.2 c 14.1 c - 13.2 - 10.6 a - 10.0 - - 42.5 - - 24.7 ab 15.7 b Income group Frequency Low Middle Upper Urban Peri-urban Rural 3.2 c 2.2 12.1 a 2.4 b 1.9 1.8 1.0 5.7 ab 15.6 3.8 12.2 b 5.1a 2.7a 10.8 15.8 - 28.0 b 5.0 27.8 a 10.9 a 4.3 a 63.6 b 9.2 6.7 13.1 36.7 10.6 a - 8.7 6.3 3.0b 11.0 10.3a - 9.4 a 4.0 21.5 c 16.1 b 5.0 11.3 58.3 10.9a - 10.0 15.0 48.8 41.9 a - - 23.4 b 16.0 ab 3.6 b 5.0 12.6 a 2.6 a - 3.5 1.3 6.0 a 12.5 3.2 13.1 a 4.7ab 2.8a 11.3 14.1 5.0 30.4 a 5.0 28.8 a 11.3a 4.6 a 69.2 ab 14.0 6.0 - 32.9 10.8 a 9.5 8.3 20.0 3.4 a 10.0 10.2 a 9.6 9.4 a - 23.1 b 18.5 a - 14.3 50.0 11.1 a 45.0 - 12.0 50.4 44.5 a 15.0 - 25.1 a 15.3 b 3.8 a 3.0 13.0 a 2.9 a 5.0 4.0 - 5.3 b 12.5 3.5 13.2 a 4.3b 2.6a 10.0 11.1 - 31.3 4.0 28.3 a 11.4 3.9 a 75.0 a 11.6 6.3 10.0 20.0 11.2 a - 10.0 - 3.3 ab 10.0 10.7 a 13.5 9.4 a 10.0 24.0a 17.2 ab - 15.5 55.0 11.0 a - - - 51.4 45.6 a - 25.0 26.8 a 18.3 a 740 2 119 57 1 7 1 23 0 32 299 39 77 5 8 1 21 4 51 11 19 31 3 2 1 12 150 1 6 1 53 4 35 2 13 1 562 85 0 3 8 7 0 0 1 24 96 0 0 112 14 746 1 57 37 1 2 0 23 1 6 233 41 49 9 1 0 18 8 37 10 16 16 5 2 2 6 52 1 6 0 40 1 36 1 15 1 460 46 1 2 2 16 1 0 1 4 24 1 1 79 3 899 3 19 31 2 1 2 41 6 2 120 139 33 1 0 0 2 8 6 19 10 11 4 2 1 0 7 0 3 1 18 2 51 1 27 0 471 40 0 9 0 53 0 1 0 0 6 0 0 42 34 Frequency Low Middle Upper 1000 4 67 50 2 3 2 23 4 9 241 90 60 9 2 0 16 16 38 14 12 13 4 1 3 7 58 0 5 1 38 2 46 0 16 1 563 60 1 5 2 20 0 1 1 8 27 0 0 73 16 980 1 93 57 0 4 1 47 1 22 279 77 67 4 4 1 17 1 46 19 23 34 5 4 0 9 109 2 6 1 50 3 50 2 26 0 643 72 0 7 5 29 1 0 1 13 67 1 0 107 23 405 1 35 18 2 3 0 17 2 9 132 52 32 2 3 0 8 3 10 7 10 11 3 1 1 2 42 0 4 0 23 2 26 2 13 1 287 39 0 2 3 27 0 0 0 7 32 0 1 53 12
Rice 4.4 a Corn/maize 3.5 Bread 12.9 a Potato 2.9 a Sweet potato 5.0 Taro 3.8 Cassava 1.3 Glutinous rice 6.9 a Instant noodle Fresh rice noodles 3.5 Shrimp instant noodle 13.6 a Arrowroot noodles 5.9 a Rice vermicelli 2.8 a Square glutin. rice cake 11.0 Dumpling 13.3 Kind of rice-cake 5.0 Pork-pie 30.7 a Rice flour 5.3 Meat pie 29.4 a Sesame and salt 12.2 a Steam roll rice pancake 5.3 a Salted shredded meat 72.4 a Kind of fried cake 12.2 Glutinous rice cake 6.7 Rice sheet, rice paper 13.3 Instant soup 32.5 Steamed glutinous rice 10.9 a Floating cake 9.0 Pyramidal rice dumpling 10.0 Other cereals 20.0 Mungbean 3.5 a Black gram 10.0 Peanuts 10.9 a Sesame seed 13.1 Green peas 10.1 a Other pulses 10.0 Pork 26.0 a Bones of pig 19.2 a Blood of pig - Liver of pig 15.0 Kidney of pig 54.0 Insides of a pig 13.7 Tongue of pig - Puddings of pig - Feet/leg of pig 12.0 Beef (thigh) 49.2 Beef (other parts) 45.0 a Stomach of cow - Cow hooves - Chicken 26.1 a Duck 17.9 a
Appendixes
Appendix 5 (continued)
Food name Location Urban Peri-urban Rural - 21.0 a 18.0 - - 28.8 33.5 34.2 11.6 b 14.0 b 17.0 30.4 a 30.0 28.5 20.9 2.6 4.8 5.0 8.0 9.4 a 2.1 a 6.2 a 6.0 9.0 b 4.5 3.7 3.0 7.2 a 6.8 8.7 2.6 a - 5.1 b 21.0 4.0 4.8 b - 26.5 - 3.9 4.6 a 2.7 8.1 5.6 a 4.4 a 2.0 - 1.5 - - 12.0 7.0 a - - 19.5 - - - 33.7 - 14.3 8.5 c 14.8 12.6 19.5 b 30.0 40.0 12.7 b 1.8 8.8 - 8.4 6.4 b 1.9 a 5.1 5.2 8.3 - 3.2 2.2 b 6.8 a 7.2 5.0 2.0 b - 4.4 c 12.0 2.2 3.3 c - 2.2 2.0 3.5 5.6 2.8 - 5.8 3.2 b 1.5 2.3 - 4.0 4.0 - 6.9 a 3.0 Income group Frequency Low Middle Upper Urban Peri-urban Rural 40.0 20.0 18.0 33.3 - 30.0 45.8 18.9 10.1 c 13.6 b 14.5 a 27.6 a 30.0 31.7 15.4 a 2.0 6.3 - 7.1 8.5 b 2.1 a 6.6 a 6.6 9.7 b 4.0 2.6 3.1 a 7.3 a 7.1 3.7 2.5 a - 5.1 b 19.3 2.8 4.9 a - 10.3 - 3.6 4.5 ab 2.3 7.3 5.5 a 3.4 a 1.6 2.3 2.7 4.5 - 12.0 7.3 a - - - 22.9 17.8 - - - 19.3 50.0 - 40.8 30.9 32.0 40.0 46.0 16.7 11.7 b 13.7 a 17.9 ab 20.8 a 16.8 a 21.7 30.3 a 27.9 a 92.9 132.2 48.5 130.0 16.0 a 17.5 a 2.3 3.2 8.7 8.0 5.0 10.0 12.5 11.4 10.4 a 10.3 ab 2.1 a 2.1 a 6.5 a 6.2 a 6.3 8.0 11.2 a 10.1 b 5.0 - 4.9 4.6 3.4 a 2.4 7.4 a 7.4 a 7.8 8.5 6.0 4.5 2.2 a 2.5 - 20.0 5.4 a 5.2 ab 21.4 15.5 2.6 2.9 5.1 a 5.1 a 15.0 18.3 3.5 - - 2.0 4.0 5.0 4.3 b 5.2 a 3.3 3.7 8.9 a 8.3 a 5.2 a 5.5 3.8 a 3.1 a 2.5 2.2 - - 2.2 2.0 7.0 - - 4.0 - 15.0 7.3 a 7.1 a 7.0 3.0 1 16 0 4 1 3 8 0 171 58 7 79 9 5 30 4 10 1 17 107 168 89 12 78 0 7 21 86 7 11 5 1 161 14 10 121 7 1 0 2 29 13 26 18 67 5 0 6 2 1 1 76 2 0 14 1 0 0 4 2 4 133 25 15 34 1 8 9 7 5 1 5 40 78 24 9 25 2 6 5 28 4 3 16 0 143 2 1 45 0 2 0 20 17 8 7 10 42 2 0 1 0 0 1 44 0 0 3 0 0 0 7 0 2 263 5 9 43 1 2 22 5 3 0 5 16 84 6 5 2 0 10 16 13 3 2 12 0 110 1 5 15 0 5 1 2 5 4 0 2 76 11 2 0 1 1 0 14 1 Frequency Low Middle Upper 1 9 1 1 0 9 6 3 212 30 14 50 1 6 18 9 8 0 8 36 113 31 8 24 1 6 13 34 4 2 15 0 163 3 4 49 0 6 0 14 13 9 6 10 70 8 2 3 2 0 1 49 0 0 18 0 0 1 4 1 2 251 39 11 72 7 8 27 4 7 1 10 83 141 52 14 46 1 13 23 55 5 10 11 0 184 10 8 89 4 2 0 8 23 13 17 15 81 5 0 3 1 0 0 53 2 0 6 0 3 0 1 3 1 104 19 6 34 3 1 16 3 3 1 9 44 76 36 4 35 0 4 6 38 5 4 7 1 67 4 4 43 3 0 1 2 15 3 10 5 34 5 0 1 0 2 1 32 1
Goose 40.0 Swan 21.6 a Pigeon - Other bird 22.8 Leg/feet of dog 50.0 Dog (other parts) 37.8 Processed meats 45.0 Other meats - Fresh fish 15.9 a Sea fish 18.6 a Dried/processed fish 21.2 Fresh shrimps 33.4 a Sea shrimps 113.0 Dried/processed shrimp 80.0 Fresh crabs 17.4 a Fresh shells 2.6 Snail 8.5 Other water foods 10.0 Mango 11.9 Orange 10.6 a Banana 2.2 a Apple 6.6 a Sapodilla 7.8 Longan 11.0 a Plum - Pumelo 6.2 Papaya 4.0 a Mandarin 7.5 a Persimmon 8.8 Jackfruit 4.5 Starfruit 2.9 Custard apple 20.0 Lemon 5.9 a Grape 20.0 Pineapple 2.9 Watermelon 5.4 a Muskmelon 16.4 Pear-shaped melon 5.0 Casaba melon - Coconut 3.8 Dragonfruit 4.4 a Guava 3.2 Pear 8.5 Rambutan 5.2 a Kumquat 3.3 b Rose-apple 3.1 Areca - Pachyrrhizus 2.5 Momordica 6.0 Oleaster 4.0 Star apple 15.0 Dracontomelum 7.5 a Wampee 7.0
0
An Analysis of Food Demand Patterns in Hanoi
Appendix 5 (continued)
Food name Location Urban Peri-urban Rural Income group Frequency Low Middle Upper Urban Peri-urban Rural 1.7 b 2.6 a 5.8 10.5 a 14.8 2.2 a 11.3 2.5 a 2.1 a 2.0 a 1.7 2.7 3.8 1.8 a 2.0 a 1.8 4.4 2.0 a - 11.7 a 2.0 2.9 a 4.0 1.5 a 3.1 a 2.7 1.9 ab 6.4 ab 2.1 a 2.8 a 2.5 2.1 2.7 a 2.9 4.5 54.4 2.1 b 3.3 2.2 a 2.8 a 2.0 5.3 a 7.8 a 1.4 3.9 a 3.9 a 4.3 3.0 3.0 2.0 2.1 a 3.6 3.8 1.8 a 2.8 a 6.7 9.7 a 10.0 2.0 11.7 2.5 a 2.3 a 1.9 2.0 3.0 2.3 1.9 a 2.1 a 1.9 3.9 2.0 a - 12.2 a 4.0 2.9 a 4.8 1.4 a 3.1 a 3.3 2.0 a 5.3 b 2.0 2.9 a 2.5 2.3 2.6 a 3.7 4.1 - 2.5 a 4.0 2.0 a 2.7 ab - 5.1 a 7.2 ab 1.7 6.0 3.4 4.0 3.3 4.3 4.0 2.6 a 2.0 3.7 376 277 17 74 5 14 12 53 29 9 1 25 9 182 72 6 16 90 0 55 1 17 17 5 44 10 28 40 10 251 0 7 70 8 19 7 98 10 29 58 0 27 25 3 11 24 0 10 11 2 19 2 8 391 187 3 60 1 11 2 52 16 24 1 4 1 168 45 5 0 110 0 16 7 18 1 7 14 11 41 31 12 189 0 1 33 4 2 0 101 1 17 22 1 27 22 19 20 14 5 1 1 0 12 4 1 505 251 2 76 1 10 0 22 28 14 2 3 0 127 107 6 1 40 1 47 2 19 0 34 28 3 63 12 1 137 3 8 131 6 7 0 117 2 17 12 0 17 62 9 20 13 8 6 1 0 29 1 3 Frequency Low Middle Upper 563 248 4 82 2 13 2 35 21 28 0 9 2 186 90 6 2 103 1 39 5 27 4 19 30 9 51 32 10 207 1 6 90 6 6 1 123 4 19 35 0 27 44 24 20 18 5 3 5 0 18 2 5 495 323 10 87 4 19 9 62 39 10 3 15 7 201 92 4 10 98 0 55 3 15 8 17 34 12 63 31 12 262 1 5 104 4 15 6 143 6 28 45 1 28 47 6 24 25 7 9 5 1 27 4 5 214 144 8 41 1 3 3 30 13 9 1 8 1 90 42 7 5 39 0 24 2 12 6 10 22 3 18 20 1 108 1 5 40 8 7 0 50 3 16 12 0 16 18 1 7 8 1 5 3 1 15 1 2
Kangkong 2.3 a Spring onion 2.9 a Bulb onion 6.0 Garlic 11.0 a Leek 13.0 Amaranth 2.2 a Celery 11.4 Ceylon spinach 2.8 a Wax gourd 2.9 a Bottle gourd 2.1 Bitter gourd 1.5 Cucumber 2.8 Chinese cabbage 3.4 Common cabbage 2.2 a Vegetable mustard 2.5 a Turnips 2.1 Cauliflower 4.4 Pak-choi 2.3 a Sweet pepper - Hot chilies 11.4 a Pumpkin 4.0 Pumpkin buds 3.0 a Carrot 4.5 Radish 1.8 Lettuce 3.9 a Eggplant 3.0 a Smooth loofah 2.5 a Ginger 6.3 a Chayote 2.3 a Tomato 3.1 a Flower of banana - Yardlong bean 2.2 Star gooseberry 3.7 a Coriander 3.1 Salad 5.1 Mushroom 58.6 Kohlrabi 2.9 a Tamarind 4.2 Garland chrysanthemum 2.4 a French bean 2.9 a Leafy vegetable - Bamboo shoot 6.2 a Spring onion bulb 8.7 a Sweet potato buds 2.1 Other vegetable 5.2 a Fennel 4.0 a Persicaria - Basil 3.9 Perilla 3.8 Marjoram 3.0 Water dropwort 3.3 a Lemon grass 2.3 Eryngium 4.0
1.5 b 1.2 c 1.5 c 2.4 b 2.5 b 2.6 a 6.1 7.5 6.0 9.5 b 9.5 b 9.7a 10.0 10.0 7.8 2.1 a 2.0 a 2.0 a 8.3 - 8.3 2.3 b 1.8 c 2.1 b 2.1 b 1.4 c 2.1 a 1.8 a 1.8 a 1.8 a 1.7 2.0 - 2.8 1.6 2.4 3.3 - 2.3 1.5 b 1.1 c 1.4 b 1.9 b 1.5 c 1.8 b 1.7 1.5 1.7 2.0 4.3 1.7 b 1.4 c 1.8 b - 2.0 2.0 10.9 a 10.2 a 8.9 b 2.4 1.8 2.1 2.6 a 2.0 b 2.2 b 2.0 - 4.5 1.7 1.4 1.5 a 2.6 b 1.9 c 2.9 a 2.3 a 1.1 1.8 1.9 b 1.5 c 1.7b 6.3 a 6.3 a 6.8 a 1.9 a 1.0 1.9a 2.6 b 2.3 c 2.5 b - 2.3 2.0 2.1 1.8 1.7 2.9 b 2.0 c 2.6 a 3.4 3.8 3.3 3.4 2.8 4.4 - - 83.3 2.0 b 1.3 c 1.7c 4.0 2.0 4.5 2.2 a 1.3 b 1.9 a 2.3 b 2.0 b 2.4 b 2.0 - - 5.2 b 3.8 c 5.3a 5.9 b 6.9 b 6.4b 1.1 0.9 1.0 3.1 b 5.4 a 4.6 a 4.1 a 3.7 a 4.4 a 6.2 4.6 6.8 2.5 2.6 4.4 2.0 2.0 3.6 - - - 2.0 b 1.6 c 2.2a 3.5 5.0 3.5 2.0 3.3 3.5
Appendixes
Appendix 5 (continued)
Food name Pot-herbs Indian taro Cress Baby corn Cow milk Other milk Powder milk Hen egg Duck egg Other egg Black tea Green tea Coffee Sugar Butter Vinegar Pickles Juices Soda (Coca-cola, etc.) Iodized salt Glutamate Other salt Cooking oil Pig fat Sweet cake Soybean cake Biscuits Fish sauce Soybean sauce Black pepper Soya-sauce Beers Other alcohols Sugarcane Eugenia Galingale Shrimp paste Candy Other foods Silkworm cocoon Saffron Chutney Artemisia Ferment Cat’s ear Meat roll Sausage Fried fish Absinth Canned meat Pie Chinese sausage Soybean milk flour Lotus seed Location Urban Peri-urban Rural Income group Frequency Low Middle Upper Urban Peri-urban Rural 2.7 a - 2.8 - 16.5 a 19.5 95.9 a 17.8a 14.7 a 22.5 40.0 31.3 a 83.8 a 5.8 a 50.0 3.0 a 2.7 a 20.0 7.6 8.0 a 25.2 a 1.4 a 14.6 a 10.6 a 25.9 a 3.6 a 40.0 9.1 a 3.4 a 50.0 10.0 10.1 a 7.3 a 2.2 17.8 4.9 19.6 - 25.3 ab 20.0 10.0 12.5 15.0 - 20.0 - 30.0 20.0 - - 21.7 - 25.0 18 1 7 1 84 5 39 135 145 4 8 181 33 88 3 25 129 0 4 722 479 70 481 131 35 324 1 652 4 5 2 61 114 4 2 9 11 0 110 4 4 2 2 0 1 3 3 3 1 3 8 2 6 1 18 3 0 0 21 5 13 75 159 0 1 245 5 51 0 12 151 0 4 563 654 245 177 424 9 332 2 672 47 2 1 32 200 5 4 20 7 0 39 6 2 0 2 2 2 3 0 0 0 1 3 0 5 0 38 2 0 0 9 0 21 99 148 3 0 416 1 77 0 26 247 1 2 414 831 559 60 613 1 277 0 799 115 4 1 36 300 13 24 16 15 1 50 7 15 1 2 3 3 1 0 0 1 0 0 0 3 0 Frequency Low Middle Upper 21 3 1 1 29 2 22 106 187 1 0 350 7 84 0 25 228 0 2 654 878 412 199 591 8 400 1 884 78 1 0 37 253 8 7 25 16 1 58 10 6 0 4 3 3 5 0 2 0 2 2 0 6 0 42 3 3 0 56 5 31 146 186 4 8 346 15 96 1 23 224 0 7 729 795 333 357 428 20 390 1 880 62 9 3 58 263 8 16 15 14 0 92 6 12 2 1 2 1 2 1 0 2 2 6 2 8 0 11 0 3 0 29 3 20 57 79 2 1 146 17 36 2 15 75 1 1 316 291 129 162 149 17 143 1 359 26 1 1 34 98 6 7 5 3 0 49 1 3 1 1 0 2 0 2 1 0 0 3 0 0 1
3.4 a 2.3 b 2.0 c 2.0 b 2.4 a 3.0 2.6 2.7 2.6 2.8 2.5 - - 2.0 2.5 10.0 - - 10.0 - 15.3 a 15.3 a 15.9 14.5 b 15.2 ab 21.8 13.8 - 15.0 17.9 94.5 a 104.3 a 100.1 a 98.7 a 98.5 a 17.1 a 15.8 b 17.3 a 16.2 b 16.8 b 15.1 a 14.2 b 14.2 b 14.2 b 14.7 a 20.0 - 18.9 16.7 18.8 50.5 30.0 - - 49.2 35.4 a 30.9 b 27.5 c 29.3 b 30.7 a 92.4 80.0 100.0 103.6 93.3 a 6.0 a 5.6 b 5.2 c 5.5 a 5.7 a 43.8 - - - 31.3 3.3 a 2.5 b 2.4 b 2.4 a 3.1 a 2.9 a 2.4 b 2.0 c 2.2c 2.4 b - - 20.0 - - 10.2 7.4 7.1 7.6 8.8 7.6 b 7.7 b 8.0 a 7.6 b 7.7 b 25.2 a 25.1 b 25.1 b 25.2 a 25.1 a 1.3 b 1.4 ab 1.4 a 1.4 a 1.4 a 14.4 a 14.2 b 13.7 c 14.0 c 14.4 b 10.4 a 10.1 b 10.7 a 10.3 b 10.6 a 25.9 20.4 20.0 21.0 25.0 a 3.9 a 3.6 b 3.4 c 3.6 a 3.6 a 40.0 50.0 - 50.0 50.0 11.3 a 6.0 b 4.3 c 5.6 c 7.6 b 3.9 3.8 a 3.9 a 4.1 a 3.7 a 80.0 50.0 67.5 50.0 74.4 10.0 10.0 10.0 - 10.0 11.1 a 6.3 b 4.7 c 6.4 b 8.1 a 8.1 a 6.5 b 5.6 c 6.0c 6.4 b 2.3 2.8 2.7 2.2 3.5 5.0 9.8 16.0 3.9 17.6 4.9 3.7 a 5.1 a 4.0 a 5.0 a 14.4 a 9.4 11.4 a 9.9 a 12.7 a - - 5.0 5.0 - 27.7 a 20.9 b 18.3 b 21.2 b 25.2 a 18.0 13.2 6.1 10.6 11.3 11.3 12.5 6.2 9.4 6.4 13.5 - 12.5 - 13.5 13.5 22.0 15.0 17.8 15.0 - 2.0 2.8 2.0 3.2 25.0 26.7 25.0 29.4 25.0 23.3 15.0 15.0 16.0 25.0 30.0 - - - 30.0 20.7 21.0 5.0 - 3.0 - 4.0 40.8 40.0 - 41.3 40.0 21.3 10.0 - 15.0 17.5 30.0 - - - 30.0 1.7 1.5 1.4 1.7 1.4 25.0 - - - -
An Analysis of Food Demand Patterns in Hanoi
Appendix 6
Demand elasticities of major food groups by location and income group in Hanoi
Food group Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Percentage change in the quantity consumed with one percentage change in the price of: Aquatic Egg and Cereals Vegetables Fruits Meats products milk Others Income –0.0894 –0.0087 –0.0114 –0.0208 –0.0073 –0.0054 –0.0060 –0.0128 –0.1490 –0.0066 –0.0121 –0.0043 –0.0032 –0.0035 –0.1398 –0.0555 –0.2525 –0.1324 –0.0466 –0.0347 –0.0382 –0.3707 –0.1161 –0.6830 0.0836 –0.0058 –0.0037 –0.0315 –0.8161 –0.2266 –1.0253 0.4438 –0.0727 0.0351 –0.0268 Urban –0.3176 –0.1260 –0.1649 –0.7571 –0.1058 –0.0788 –0.0868 Peri-urban –0.4274 –0.1339 –0.0355 –1.0779 –0.0067 –0.0042 –0.0363 Rural –0.4396 –0.1220 0.0013 –1.1931 –0.0392 0.0189 –0.0144 –0.2654 –0.1053 –0.1378 –0.2513 –0.4100 –0.0658 –0.0725 –0.1988 –0.0789 –0.1032 –0.1882 –0.0662 –0.3683 –0.0543 –0.2480 –0.0984 –0.1288 –0.2348 –0.0826 –0.0615 –0.5336 –0.2581 –0.0808 –0.0214 0.0582 –0.0041 –0.0026 –0.8493 –0.2343 –0.0650 0.0007 0.1274 –0.0209 0.0101 –0.9471 –0.2609 –0.0761 –0.0132 0.0894 –0.0056 0.0029 –0.8790 –0.2445 –0.0822 –0.0525 –0.0383 –0.0433 –0.0289 –0.7578 0.1098 0.0654 0.7665 1.6605 1.3919 1.0950 1.3147 0.0792 0.0599 1.7164 1.7775 1.7663 1.3878 1.0077 0.0743 0.0634 3.5566 1.8827 1.3535 1.4351 0.9435 0.0693 0.0576 2.4316 1.9739 1.8321 1.5242 0.9505 0.0906 0.0647 1.2761 1.7217 1.3893 1.2377 1.0965
Cereals –0.1389 –0.0153 Vegetables –0.0065 –-0.2008 Fruits –-0.0017 –0.0013 Meats 0.0047 0.0035 Aquatic products –0.0003 –0.0002 Egg and milk –0.0002 –0.0002 Others –0.0018 –0.0013 Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others Cereals Vegetables Fruits Meats Aquatic products Egg and milk Others –0.2070 –0.0053 0.0001 0.0104 –0.0017 0.0008 –0.0006 –0.1749 –0.0058 –0.0010 0.0068 –0.0004 0.0002 –0.0013 –0.1284 –0.0071 –0.0045 –0.0033 –0.0037 –0.0025 –0.0030 –0.0161 –0.2626 0.0001 0.0088 –0.0014 0.0007 –0.0005 -0.0161 -0.2190 -0.0008 0.0055 -0.0004 0.0002 -0.0011 –0.0141 –0.2132 –0.0030 –0.0022 –0.0025 –0.0017 –0.0020
–0.4217 –0.3463 –0.1321 –0.1085 –0.0350 –0.0288 0.0951 0.0781 –0.9558 –-0.0054 –0.0042 –-0.9561 –0.0358 –0.0294 –0.3363 –0.0934 0.0010 0.1829 –0.8000 0.0145 –0.0110 –0.5012 –0.1461 –0.0254 0.1717 –0.9283 0.0057 –0.0338 –0.2963 –0.0996 –0.0636 –0.0464 –0.6495 –0.0350 –0.0426 –0.3417 –0.0949 0.0010 0.1858 –0.0304 –1.1881 –0.0112 –0.4152 –0.1210 –0.0210 0.1423 –0.0089 –1.0600 –0.0280 –0.2453 –0.0824 –0.0527 –0.0384 –0.0435 –0.6291 –0.0353
Low income –0.5469 –0.5157 –0.1594 –0.1503 –0.7148 –0.0261 0.1874 –1.1428 –0.0117 –0.0111 0.0062 0.0058 –0.0368 –0.0347 High income –0.2184 –0.3556 –0.0734 –0.1195 –0.5169 –0.0764 –0.0342 –0.9550 –0.0387 –0.0630 –0.0258 –0.0420 –0.0314 –0.0512