Using the receipts from shops and supermarkets collected as part of a Household Budget Survey to estimate Food consumption
July 2006
Kevin McCormack Central Statistics Office Ireland
1
1.
Introduction
The Irish Household Budget Survey (HBS) is a survey of a representative random sample of all private households in the State. It has been undertaken at regular intervals since 1951 and on a five-yearly basis since 1994. The main purpose of the HBS is to determine in detail the current pattern of household expenditure in order to update the weighting basis of the Consumer Price Index. The maintenance of a detailed diary of household expenditure over a twoweek period by the surveyed households is thus the main distinguishing feature of the HBS. In addition, detailed information on all sources of household income is also collected as part of the survey details. In order to reduce the burden on households when completing the detailed expenditure diaries households have been encouraged since 1994 to attach till receipts (scanner receipts) that contain the details of the purchases to their diaries instead of directly recording such information. Analysis shows that the vast majority of the scanner receipts relate to the purchase of food items in shops and supermarkets (i.e. food outlets). These receipts contain very detailed descriptions of the individual food items purchased as well as their prices and associated weights (grams) or volumes (litres). The scanner receipts also aid the CSO when processing data collected from the surveyed households as the item descriptions and price information allow for more accurate coding by reducing transcription errors that can occur when respondents are completing their diaries by hand. Currently when processing the data on food items only the item descriptions (in the form of a 3-digit identity code) and prices on the scanner receipts are used. This leaves an important source of information on food, namely the data on weights and volumes, unused. This study makes use of the weight and volume data from a subsample of the households surveyed in the HBS in order to develop a method which allows for the conversion of the published average weekly household expenditure on food to its equivalent weight or volume.
2.
Data source
The most recent survey, for which results are available, was carried out during the period June 1999 to July 2000. Some 7,644 household participated in this survey and the detailed scanner receipts data collected from a sub-sample of 945 of these households were used in this study. (See Section 4 for more details on the subsample and Appendix I for a description of the HBS sample design).
3.
Methodology
In an ideal world the weight or volume and price data for the food items purchased by responding households in a HBS should also be recorded. However, this is not normally done, as it would substantially add to the amount of time required to process the data.
2
It is possible, however, to make use of the weight and volume data without compromising data processing time-lines by: 1. selecting a representative sub-sample of responding households and recording the full item description, the product code, price and weight or volume information for the food items listed in their expenditure diaries or on attached scanner receipts 2. from the price and weight or volume data in stage 1 above, estimating the weight or volume per €1 of expenditure for each food item through the use of derived conversion factors 3. applying these conversion factors to the HBS published results to convert the expenditure data to their equivalent or corresponding weight or volume.
4.
Sub-sample
For this study the item description, price and weights or volume information on the food items listed as purchased by the households were extracted from the expenditure diaries of a randomly selected sub-sample of 945 household that participated in the 1999-2000 HBS. Thus this sub-sample is a representative sample of all households which were included in the 1999/2000 HBS. Some 70,533 individual entries relating to 121 food items were manually extracted and a conversion factor for each food item was then estimated. 4.1 Dataset
By following the stages as outlined in Section 3 above a data-set was created from the sub-sample that contains for each food item the following details: an exact product description, a 3-digit identity code and price and weight or volume. Table 1 below shows an extract from the sub-sample dataset on completion of Stage 1 of the procedure. There are 10 households and 18 entries for White Bread in this example. A household is identified by its area, household number (HLD) and population density stratum references (Strata). The data was initially keyed into 4 Excel Files (it took 40 man weeks to complete this task) and then converted into a SAS dataset for final processing and analysis. Using SAS, the dataset can be organised or resorted at a household or food item level. Organising the dataset by food item allows missing data to be quickly estimated. For example, if, in Table 1 below, the 3-digit code for white bread was missing from an entry: this can quickly be identified and the correct code inserted. Another example is where an item has been miss-coded: this can identified very quickly when the dataset is organised by food item. It would be very clear that an entry for white bread has been wrongly coded if its code is 102 and not 101. A final advantage of organising the dataset by food item is that one can undertake the above changes manually.
3
Table 1: Extract from Sub-Sample data-set – HBS 1999-2000
Area HLD Strata Item Description CODE WEIGHT (G) Volume PRICE (L) (€)
9935 9935 1621 1362 1362 1362 2148 1385 9383 9383 1306 1346 1496 1550 1550 1550 1550 1550
1 1 1 2 2 2 1 3 1 1 3 3 2 2 2 2 2 2
2 2 7 5 5 5 2 7 1 1 8 7 6 4 4 4 4 4
White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread White Bread
101 101 101 101 101 101 101 101 101 101 101 101 101 101 101 101 101 101
400 400 800 800 800 800 800 800 800 800 800 800 800 800 800 800 800 800
0.41 0.41 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62 0.62
4.2
Sub-sample details
The distribution of households and the number of food entries listed in their expenditure diaries or on scanner receipts attached are presented in Table 2 below. Table 2: Sub-Sample details - HBS 1999-2000
Strata Number of Households % Number of food entries %
1. County Borough 2. Suburbs of County Boroughs 3. Environs of County Boroughs 4. Towns 10,000+ 5. Towns 5,001-10,000 6. Towns 1,000- 5,000 Total Urban Areas 7. Mixed Urban/Rural Areas 8. Rural Areas Total Rural Areas Total
126 112 11 167 20 73 509 186 250 436 945
13.33 11.85 1.16 17.67 2.12 7.72 53.85 19.69 26.46 46.15 100.00
9,513 8,503 734 11,553 1,709 4,734 36,746 13,705 20,082 33,787 70,533
13.49 12.06 1.04 16.38 2.42 6.71 52.10 19.43 28.47 47.90 100.00
Slightly over half the sub-sample of households and the number of food entries are associated with urban areas. The area distribution of sub-sample household is consistent with the parent HBS sample and as a result data for the obtained from the sub-sample can be applied with a high level of confidence to the parent sample.
4
4.3
Aggregated sub-sample measurement and expenditure values
The item descriptions, expenditures, weights, volume and number of entries recorded from the household diaries or scanner receipts for each item is presented in aggregated form in Table 3 below. Food items which are measured by weight account for 86% of the entries recorded, while food items that are measured by volume accounted for 12%. Entries for Vegetables, Meat and Bread, Flour, Biscuits & Cakes, were almost equal in number (approximately 10,000) and in total represented more than half of the entries for food items that are measured by weight. In terms of expenditure, food items measured by weight again accounted for 86% of the total with Meat showing the highest individual expenditure, while in the case of food items measured by volume Milk & Cream had the highest expenditure. Table 3: Aggregated Expenditure and Measurement Values from a subsample of the HBS 1999/2000
Item Description Grams Litres Quantity Expenditure (€) No. of entries
Food (by weight) Bread, Flour, Biscuits & Cakes Solid Milk Products Cheese Butter & Fats Meat Fish Vegetables Fruits & Nuts Miscellaneous Foods Total Food (by weight) Eggs Food (by volume) Fresh Milk & Cream Other Fats & Cooking Oils Ice Cream & Juices Soft Drinks Total Food (by volume)) Total
6,261,571 1,275,831 502,174 1,187,113 4,614,541 459,681 9,097,129 3,498,272 6,135,326 33,031,637 0
0 0 0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0 0 9,900
17,194.12 4,452.12 3,530.46 8,338.78 30,179.30 3,870.73 13,312.17 7,097.01 32,408.14 120,382.85 1,343.46
10,760 2,417 1,786 2,410 9,755 1,412 10,624 4,943 16,832 60,939 801
0 0 0 0 33,031,637
5,840.71 403.35 14,285.01 12,774.91 33,303.98 33,303.98
0 0 0 0 0 9,900
7,182.17 792.23 5,260.66 5,154.30 18,389.35 140,115.66
3,239 336 2,314 2,904 8,793 70,533
4.5
Conversion Factors
The calculation process employed to generate the conversion factors which are used in this study to convert published expenditure data for the 121 food items from the HBS 1999-2000 to their equivalent weight or values is shown in Table A in Annex II. The approach used to generate these factors is relatively straightforward.
5
Firstly, the food items are grouped together by 3-digit product identity code, which gives the total number of entries (or quotations) for each item, the total weight, volume or number recorded for each food item and the total expenditure in IRE£ recorded for each item. Secondly, the IR£s are converted to Euros € by applying the a factor of IR£1 = €1.27. Finally, the conversion factors are calculated by dividing the weight, volume or number value for each food item by the corresponding expenditure. For, example the conversion factor for white bread is calculated as 3,095,370 / €4,675.63 = 662.02 g/€1 (i.e. €1 of expenditure is the equivalent to 662.02 grams of white bread) It should be noted that there are a number of food items for which there are a very small number of entries recorded. For example, condensed milk has only three entries. As a consequence caution should be exercised when using conversion factors in those cases where there are low numbers of price observations.
5. Converting HBS expenditure to weight and volume values
Stage 3 of the approach used in this study is to apply the derived conversion factors to the published HBS results in order to convert the expenditure data to their equivalent weight or volume. 5.1 Urban/Rural Food Expenditure & Consumption
The conversion of the published expenditure classified by Urban/Rural Location is shown in detail in Table B in Annex III and in aggregated form in Table 4 below. For example, in 1999/2000 households in Urban Locations spent on average of €2.71 each week on white bread and conversion factor for white bread is €1 = 662.02 grams. Therefore, the estimated weight of white bread purchased corresponding to an expenditure of €2.71 using the above conversion factor is 1,791 grams. (See Table B1) Farm households in rural areas spent the most on the food items Farm households in rural areas spent the most on the food items in this study, at €112.55, while households in urban areas spent the least, with an average weekly expenditure of just €85.69. Farm households in rural areas also purchased the highest amount of food measured by weight (26.6kg) and by volume (28.4 litres) and households in urban areas the least, with an average weekly purchase of food measured by weight of 19.3kg and by volume of 24.5 litres. Farm households in rural areas purchased substantially more meat and vegetables than other households, at 4.6kg and 8kg respectively. (See Table 4).
1
The conversions were undertaken with un-rounded data and the results are rounded to 0 or 2 decimal places. Therefore, replication of the conversion by hand will yield slightly different results. For example, the hand calculation for white bread is 2.71 * 662.02g = 1,794g which is slightly different to the 1,791g produced from un-rounded data.
6
Chart 1. Average weekly household expenditure on Food by Urban/Rural locations - HBS 1999-2000
State
Location
Rural Non-Farm
Food by Weight Eggs Food by Volume
Rural Farm
Urban Areas
0
20
40
60 Expenditure in €
80
100
120
Vegetables accounted for slightly more than 30% of all food purchased that is measured by weight Vegetables account for slightly more than 30% of all food purchased that is measured by weight, while bread, flour, biscuits & cakes account for between 20%22% in rural households and slightly over 18% in urban households. Potatoes were the most popular choice of the vegetable, with farm households in rural areas purchasing 4.2kg each week compared with 3kg in urban areas. (See Tables 4, 5 & B) Fruits & Nuts accounted for 10% of all food purchased by weight Fruits & Nuts accounted for 10% of all food purchased by weight. Apples, oranges and bananas were the most popular of all fruits purchased with farm households in rural areas purchasing 1.8kg per week and households in urban areas purchasing 1.4kg per week. (See Tables 4, 5 & B) Less than 400g of fish purchased each week Between 250g and 320g of fish was purchased each week by the households in the different locations and this accounts for slightly over 1% of all the food purchased by weight. Fresh and tinned fish were the most popular types purchased. (See Tables 4,5 & B) Farm households purchased more tea Farm households in rural areas purchased 40% more breakfast cereals, tea and sugar in total (1.5kg) than households in urban areas (0.9kg). Households in Urban areas purchased the most pre-prepared meals, at 0.7kg) (i.e. prepared food, frozen diners, pizza & pasta) which is 30% more than households in rural locations. (See Table B)
7
Chart 2: Food purchased by weight in % by category for Households in Urban Areas - HBS 1999-2000
Miscellaneous Foods 17.2%
Bread, Flour, Biscuits & Cakes 18.6% Solid Milk Products 3.1%
Fruits & Nuts 10.7%
Cheese 1.4% Butter & Fats 1.3%
Meat 15.1% Fish 1.3%
Vegetables 31.3%
Chart 3: Food purchased by weight in % by category for Rural Farm Households - HBS 1999-2000
Miscellaneous Foods 14.2%
Bread, Flour, Biscuits & Cakes 22.29%
Fruits & Nuts 9.5% Solid Milk Products 2.25% Cheese 1% Butter & Fats 1.8% Meat 17.5% Vegetables 30.3% Fish 1.2%
8
Farm households drink more milk than their urban counterparts Some 31% of food measured by volume is accounted for by fresh milk and cream in farm households in rural areas, as compared to almost 20% for households in urban areas. Ice cream, juices and soft drinks account for almost 80% of food purchased by households in urban areas that is measured by volume. (See Tables 4 & 5) 5.2 Urban / Rural Location per person
It can be seen in Table 4 that the composition of the households in the rural/urban locations are different. Farm households in rural areas tend to be larger, with an average size of 3.56 person, while households in urban areas are the smallest with an average size of 3 person. This difference in household composition explains, to some extent, why farm households in rural areas have the highest expenditure on the food items in this study (i.e. the larger the household the greater the amount of food consumed). It is, therefore, its apparent that in order to obtain a more realistic comparison between the purchases of food by the different types of households that the weight and volume estimates presented in Table 4 must be adjusted to relate to the average consumption per person. In this way the difference in household size and composition can be reduced or eliminated. The left hand side of Table 6 below and Table C in Annex IV present the estimates of the average weekly weight or volume of food items purchased per person by households in urban and rural locations. Almost 1.5 kg more of food that is measured by weight is purchased per person in farm households than per person for households in urban areas. Almost all this difference is accounted for by bread (469g), meat (334g) and vegetables (670g). (See Tables 6 and C) The per person analysis is not a standardised approach to use when undertaking household comparisons due to varying age and sex compositions of the households and as a result no further analysis of the per person data is undertaken in this study. However, interested parties can use the left hand side of Tables 6 & C. 5.3 Urban / Rural Location per adult equivalent
On further examination of the composition of the household in the urban/rural locations it can be seen that there are differences in the age and sex distributions. To obtain more relevant comparison between the households in the different locations the age and sex distributions in these households must be taken into account. The approach used in this study is to estimate the total number of adult equivalents in each household in order to obtain a standard measurement for comparisons. In 2005 European Food Information Council (EUFIC) indicated that on average a male of age 14 years or more requires some 2,500 calories per day. A female of age 14 years or more requires 2,000 calories per day. Children under 5 require 1,300 calories per day. Male children aged 5-13 years require 1,970 calories per day. And female children aged 5-13 years require 1,740 calories per day. This information was used to estimate the number of adult equivalent persons in a household consuming
9
2,500 calories (i.e. a standard measurement). Males aged 14 years and older are given a weight of 1 (equals 2,500 calories), females aged 14 years and older a weight of 0.8, boys aged 5-13 a weight of 0.79, girls aged 5-13 a weight of 0.7 and children under 5 years a weight of 0.52. The estimated number of adult equivalents using this approach distributed by location is presented on the top of the right hand side of Table 6 below. While the centre of Tables 6 and C present the detailed and abridged food data per adult equivalent. In practice this data is the household data from Tables 4 & B divided by the number of adult equivalents in each location. Rural households purchased the most food All rural households purchase more food that is measured by weight per adult equivalent than households in urban areas (7.6kg), with farm households purchasing 1.1kg more and other rural household 0.6kg more. In terms of bread, meat and vegetables farm households purchased in total 6kg per adult equivalent each week; which is 1.1kg more than households in urban areas and 0.5kg more than other households in rural areas. (See Tables 6 & C)
Chart 4: Average weekly weight of Food purchased per adult equivalent by Urban/Rural Location - HBS 1999-2000
Miscellaneous Foods Fruits & Nuts Vegetables Fish
Food Item
Meat Butter & Fats Cheese Solid Milk Products Bread, Flour 0 500 1,000 1,500 2,000 2,500
Weight (grams)
Urban Areas
Rural Farm
Rural Non-farm
State
10
Potatoes were the most popular of the vegetables Potatoes were the most popular of the vegetables, with farm households in rural areas purchasing 1.4kg per adult equivalent each week while households in urban areas purchased 1.2kg per adult equivalent each week. (See Tables 6 & C) Apples, oranges and bananas were the most popular of the fruits Apples, oranges and bananas were the most popular of the fruits purchased with all households purchasing approximately 0.6kg per adult equivalent each week and households in urban areas purchasing 1.4kg per week. (See Tables 4, 5 & B) It should be noted that the percentage distribution of the purchases of food items per adult equivalent are the same as data presented on the right hand side of Table 5 as the estimate of the adult equivalent comparisons are simply the household data divided by a constant factor. Farm households purchased twice as much sugar as urban households Farm households in rural areas purchased almost twice as much sugar (215g) per adult equivalent each week as households in urban areas (116g). While households in urban areas purchased almost twice as much pizza (82g) per adult equivalent per week as farm households in rural locations (44g). (See Table C)
11
Table 4: Average Household Size, Weekly Household Expenditure on Food and conversion to equivalent Weight & Volume, HBS1999-2000, classified by Urban Rural Location
Item Description House Composition Males Females Total Persons Urban Areas No. 1.42 1.58 3.00 Rural Areas Farm Hlds Other Hlds All Rural Hlds No. 1.98 1.58 3.56 No. 1.57 1.6 3.16 No. 1.65 1.59 3.24 State No. 1.5 1.58 3.08 Urban Areas No. 1.42 1.58 3.00 Rural Areas Farm Hlds Other Hlds No. 1.98 1.58 3.56 No. 1.57 1.6 3.16 All Rural Hlds No. 1.65 1.59 3.24 State No. 1.5 1.58 3.08
Food (by weight) Bread, Flour, Biscuits & Cakes Solid Milk Products Cheese Butter & Fats Meat Fish Vegetables Fruits & Nuts Miscellaneous Foods Total Food (by weight)
€ 9.91 2.12 1.84 1.68 19.08 2.20 8.06 4.20 20.18 69.27
€ 14.47 2.06 1.94 3.25 31.18 2.79 10.02 4.91 20.15 90.78
€ 11.16 2.03 1.75 2.29 23.50 2.07 8.61 4.15 18.95 74.51
€ 11.81 2.04 1.79 2.48 25.04 2.21 8.86 4.28 19.22 77.74
€ 10.58 2.10 1.82 1.98 21.23 2.18 8.38 4.23 19.83 72.33
grams 3,587 598 262 247 2,917 250 6,041 2,061 3,317 19,280 No.
Weights / Volume / Number grams Grams grams 5,926 4,403 4,700 599 276 480 4,648 323 8,051 2,525 3,761 26,589 No. 573 249 333 3,553 239 6,759 2,089 3,282 21,480 No. 8.9 litres 8.86 0.24 6.12 13.20 7.2 Litres 6.24 0.24 6.65 11.55 580 255 360 3,772 255 7,008 2,161 3,379 22,469 No. 7.6
grams 3,984 594 259 291 3,226 249 6,397 2,098 3,340 20,437 No. 6.8 litres
Eggs Food (by volume) Fresh Milk & Cream Other Fats & Cooking Oils Ice Cream & Juices Soft Drinks Total Food (by volume) Total Expenditure on All Food
0.86
1.21
0.98
1.03
0.92
6.4 litres 4.83 0.21 7.19 12.31
5.92 0.42 3.09 6.13 15.56 85.69
10.80 0.47 3.16 6.13 20.56 112.55
7.61 0.47 3.07 5.80 16.95 92.44
8.26 0.47 3.09 5.88 17.69 96.46
6.75 0.43 3.09 6.03 16.30 89.55
litres 6.77 0.24 6.51 11.90
5.52 0.22 6.97 12.14
24.54
28.42
24.68
25.43
24.85
12
Table 5: Average Household Size, Weekly Household Expenditure on Food and conversion to equivalent Weight & Volume in percentage terms, HBS1999-2000, classified by Urban Rural Location
Item Description Urban Areas Rural Areas Farm Hlds Other Hlds House Composition Males Females Total Persons No. 1.42 1.58 3.00 No. 1.98 1.58 3.56 No. State All Rural Hlds No. No. 1.5 1.58 3.08 No. 1.42 1.58 3.00 Urban Areas Rural Areas Farm Hlds No. Other Hlds No. State All Rural Hlds No. No. 1.5 1.58 3.08
1.57 1.65 1.6 1.59 3.16 3.24 Expenditure as a %
1.98 1.57 1.65 1.58 1.6 1.59 3.56 3.16 3.24 Weights / Volume / Number as a % grams 22.29 2.25 1.04 1.81 17.48 1.21 30.28 9.49 14.15 100.00 No. 100.00 litres 31.18 0.84 21.52 46.46 grams 20.50 2.67 1.16 1.55 16.54 1.11 31.46 9.73 15.28 100.00 No. 100.00 litres 25.26 0.97 26.96 46.81 grams 20.92 2.58 1.13 1.60 16.79 1.13 31.19 9.62 15.04 100 No. 100.00 litres 26.62 0.94 25.62 46.82
Food (by weight) Bread, Flour, Biscuits & Cakes Solid Milk Products Cheese Butter & Fats Meat Fish Vegetables Fruits & Nuts Miscellaneous Foods Total Food (by weight)
14.30 3.06 2.66 2.42 27.54 3.17 11.64 6.07 29.13 100.00
15.93 2.27 2.14 3.58 34.35 3.08 11.04 5.41 22.20 100.00
14.98 2.73 2.35 3.07 31.53 2.78 11.56 5.57 25.43 100.00
15.19 2.63 2.30 3.19 32.22 2.84 11.40 5.51 24.72 100.00
14.63 2.91 2.52 2.73 29.35 3.02 11.59 5.85 27.42 100.00
grams 18.60 3.10 1.36 1.28 15.13 1.30 31.33 10.69 17.20 100.00 No. 100.00 Litres 19.69 0.87 29.27 50.16
grams 19.49 2.90 1.27 1.42 15.79 1.22 31.30 10.27 16.34 100.00 No. 100.00 litres 22.23 0.89 28.03 48.86
Eggs Food (by volume) Fresh Milk & Cream Other Fats & Cooking Oils Ice Cream & Juices Soft Drinks Total Food (by volume)
100.00 38.04 2.69 19.84 39.43 100.00
100.00 52.50 2.29 15.38 29.83 100.00
100.00 44.87 0.47 3.07 5.80 54.22
100.00 46.66 0.47 3.09 5.88 56.10
100.00 41.41 0.43 3.09 6.03 50.96
100.00
100.00
100.00
100.00
100.00
13
Table 6: Average Household Size, Weekly expenditure on Food per person and adult equivalent converted to equivalent Weight & Volume, HBS1999-2000, classified by Urban Rural Location
Item Description Urban Areas No. 1.42 1.58 3.00 grams 1,196 199 87 82 972 83 2,014 687 1,106 6,427 No. Eggs Food (by volume) Fresh Milk & Cream Other Fats & Cooking Oils Ice Cream & Juices Soft Drinks Total Food (by volume) litres 1.61 0.07 2.40 4.10 8.18 2.1 litres 2.49 0.07 1.72 3.71 7.98 Rural Areas Farm Hlds Other Hlds All Rural Hlds Per Person House Composition Males Females Total Persons Food (by weight) Bread, Flour, Biscuits & Cakes Solid Milk Products Cheese Butter & Fats Meat Fish Vegetables Fruits & Nuts Miscellaneous Foods Total Food (by weight) No. 1.98 1.58 3.56 grams 1,665 168 78 135 1,306 91 2,262 709 1,056 7,469 No. 2.5 litres 1.97 0.08 2.11 3.66 7.81 No. 1.57 1.6 3.16 grams 1,393 181 79 105 1,124 76 2,139 661 1,039 6,797 No. 2.3 litres 2.09 0.07 2.01 3.67 7.85 No. 1.65 1.6 3.24 grams 1,451 179 79 111 1,164 79 2,163 667 1,043 6,935 No. 2.3 litres 1.79 0.07 2.26 3.94 8.07 No. 1.5 1.58 3.08 grams 1,293 193 84 94 1,048 81 2,077 681 1,084 6,635 No. 2.2 Litres 1.90 0.08 2.83 4.85 9.66 No. 1.33 1.21 2.54 Grams 1,412 235 103 97 1,149 98 2,378 811 1,306 7,590 No. 2.5 Litres 2.88 0.08 1.99 4.29 9.23 No. 1.87 1.21 3.08 grams 1,924 194 90 156 1,509 105 2,614 820 1,221 8,632 No. 2.9 litres 2.35 0.09 2.51 4.36 9.31 State Urban Areas Rural Areas Farm Hlds Other Hlds All Rural Hlds No. 1.55 1.21 2.75 grams 1,709 211 93 131 1,372 93 2,548 786 1,229 8,171 No. 2.7 litres 2.46 0.09 2.37 4.33 9.25 2.8 litres 2.12 0.08 2.68 4.67 9.56 State
Per Adult Equivalent No. 1.45 1.20 2.65 grams 1,661 216 94 126 1,341 90 2,550 788 1,239 8,105 No. No. 1.39 1.21 2.60 grams 1,532 228 100 112 1,241 96 2,460 807 1,285 7,860 No. 2.6
Weights / Volume / Number
Weights / Volume / Number
14
5.4
Urban/Rural Location - calorific content of food purchased per adult equivalent
The detailed weight and volume food data in this study allow for the estimation of their calorific content. This can be done either at a detailed level or at an aggregated level. Table 7 below presents the estimation of the calorific content of the food items at an aggregated level per adult equivalent purchased each week by households in urban/rural locations. For each food category on the left hand side of Table 7 the average calorific value of its components were obtained per 100g or litre from international nutrition sources (i.e. 1st. column of the table). Then by simply multiplying these values by the corresponding average weekly purchases of food per adult equivalent measured by weight, No., and volume (i.e. columns 2-6 of Table 7) the corresponding calorific values are estimated (i.e. columns 7-11 of Table 7). Adults in Farm households had highest amount of calories available On converting the average purchases per adult equivalent it can be seen that the purchases that farm households in rural areas make have the highest calorific value, at 22,518 per week. This is 1,600 calories more that other rural households and over 2,900 calories more than households in urban areas. (See Table 7 & Chart 1) Soft drinks accounted for over 10% of total calories Over 10% of total calorific value from the food purchased by households in urban areas is from soft drinks, as compared to 8% in rural households. (See Table 7) Bread, butter & fats, meat and vegetables are the main sources of calories The total calorific value of bread, butter & fats, meat and vegetables purchased per adult equivalent each week by farm households in rural areas, at 13,000, calories accounts for 58% of the total and is 3,400 calories more than households in urban areas and of their calories.
15
Chart 5: Calorific value, per adult equivalent, of food purchased each week HBS 1999-2000
26,000 24,000 22,000 20,000
Calories per week
18,000 16,000 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0 Urban Areas Farm households Other Rural State
Food by volume Food by No. Food by weight
Location
6. Further Analysis – income deciles
The method to convert published HBS food expenditure data to their weight & volume equivalents as demonstrated above for household location can be readily applied to other HBS results. Tables D to I in Appendix V present, in aggregated format, the results of applying the method to the average weekly expenditure on Food data from the 1999-2000 HBS classified by gross household income groups. Tables D, E & F show the expenditure data at the household, person and adult equivalent levels, while Tables G, H & I show the weight and volume estimates after the conversion method has been applied. This paper does not intend to discuss these estimates in detail, However, some important details stand out very clearly. The average household size increases with increasing income and as the households get larger their average weekly expenditure per adult equivalent on Food measured by weight decreases. Households with the lowest income are the smallest in size (one person) and spend on average €39 per adult equivalent on Food each week: this is some €4 more than any other household (i.e. it is more expensive to buy Food for one person living on their own than for the average cost of two or more buying together). (See Table F)
16
Table 7: Average Household Size, Weight & Volume Food purchased weekly per adult equivalent converted to calorific equivalent., HBS19992000, classified by Urban Rural Location
Item Description Calories per 100g or 1L Urban Areas Rural Areas Farm Hlds Other Hlds All Rural Hlds State Urban Areas Rural Areas Farm Hlds Other Hlds All Rural Hlds Per Adult Equivalent No. No. No. 1.87 1.45 1.55 1.21 1.20 1.21 3.08 2.65 2.75 Calories 4,810 194 359 1,170 4,527 210 2,091 410 4,274 18,045 289 4,154 216 376 943 4,022 180 2,040 394 4,335 16,661 Calories 272 Calories 1,332 17 566 1,987 3,902 19,580 2,014 16 397 1,758 4,184 22,518 1,647 18 502 1,788 3,955 20,888 1,723 17 474 1,775 3,989 21,134 1,487 17 536 1,915 3,955 20,293 4,273 211 370 983 4,115 185 2,039 393 4,300 16,869 276 State
House Composition Males Females Total Persons Food (by weight) Bread, Flour, Biscuits & Cakes Solid Milk Products Cheese Butter & Fats Meat Fish Vegetables Fruits & Nuts Miscellaneous Foods Total Food (by weight) Eggs Food (by volume) Fresh Milk & Cream Other Fats & Cooking Oils Ice Cream & Juices Soft Drinks Total Food (by volume) Total Calories 700 200 200 410
No. 1.33 1.21 2.54 grams 1,412 235 103 97 1,149 98 2,378 811 1,306 7,590 No. 2.5 litres 1.90 0.08 2.83 4.85 9.66
250 100 400 750 300 200 80 50 350
100
Per Adult Equivalent No. No. No. 1.87 1.45 1.55 1.21 1.20 1.21 3.08 2.65 2.75 Weights / Volume / Number grams grams grams 1,924 1,661 1,709 194 216 211 90 94 93 156 126 131 1,509 1,341 1,372 105 90 93 2,614 2,550 2,548 820 788 786 1,221 1,239 1,229 8,632 8,105 8,171 No. No. No. 2.9 2.7 2.8 litres 2.88 0.08 1.99 4.29 9.23 litres 2.35 0.09 2.51 4.36 9.31 litres 2.46 0.09 2.37 4.33 9.25
No. 1.39 1.21 2.60 grams 1,532 228 100 112 1,241 96 2,460 807 1,285 7,860 No. 2.6 litres 2.12 0.08 2.68 4.67 9.56
No. 1.33 1.21 2.54
No. 1.39 1.21 2.60
3,530 235 412 728 3,446 197 1,903 406 4,570 15,428 251
3,831 228 398 838 3,723 191 1,968 403 4,496 16,077 261
17
7. Conclusions
The goal of this study was to expand the use of the information collected from households that participated in the Irish 1999-2000 Household Budget Survey. The detailed descriptions (including weight and volume data) on the receipts that the households received when they purchased food items were identified as important sources of additional information which add value to existing data sources. It was clear that if the weight and volume data for the food items could be processed as part of the HBS project then this data could be presented as part of the main results of the survey. To record all the weight and volume data would substantially add to the amount of time required to process the HBS and for this reason it was not undertaken for the 1999-2000 survey. However, by following an approach that: firstly selects a representative sub-sample of the households that participated in this survey: secondly records their weight and volume data and: finally uses this information estimate the weight and volume values for the main sample from the published results would expand the use of the information collected without compromising data processing time-lines. A sub-sample consisting of 945 households was selected and the detailed information on the food items they purchased was recorded. From this information conversion factors were calculated for 121 individual food items. These factors were applied to published 1999-2000 HBS expenditure results to estimate corresponding weights and volume values. Thus this study has proven that weight and volume data for food items can be estimated from a HBS in such a way that processing time-lines are not compromised. The presentation of the weight and volume data and the conversions to their calorific values also allow for expert users, such as health professional, to become more involved in using HBS results for public policy and analysis. For example, the European Commission has developed a Pan-European food data bank based on the expenditure data from household budget surveys through the DAta Food NEtworking (DAFNE) project. The aim of this initiative is to provide a nutrition monitoring tool that can assist in the formulation, implementation and evaluation of nutritional and public health policies across Europe. The specific applications of the DAFNE database are wide-ranging. One main output is the identification of patterns and trends in nutritional practices. Comparisons are conducted of the average daily individual access to foods, sorted by group, both within and between country populations. Table 8 below compares average daily individual access to food for the 1990s from the DAFNE data base to per adult equivalent data from this study for 3 comparable food items. One can note that the data are quite comparable and as result the method proposed in this study to convert the published average weekly household expenditure on food from a HBS to its equivalent weight or volume can be applied with confidence.
18
Table 8: Average daily individual access to food, comparison to DAFNE average for the 1990s Urban Rural Food Item McCormack DAFNE Diff McCormack DAFNE Diff +/+/Bread, Flour, etc. Meat Vegetables 201g 164g 339g 217g 179g 319g - 17g - 15g + 20g 244g 196g 364g 234g 178g 353g + 10g + 18g + 11g
One final point is that the detailed weight and volume data also allow CPI Statisticians to determine the most important food items to be included in their surveys.
19
Appendix I - HBS 1999-2000 – Sample design
The HBS is designed as approximately self-weighting probability sample of the noninstitutionalised Irish civilian population. This survey collects data on two different levels of analysis: household and individual. However, the sample designs are based on selecting households or dwelling units; where a dwelling unit is defined as “a room or group of rooms intended for occupation as separate living quarters and having either a separate entrance or complete cooking facilities for the exclusive use of the occupants”. Central to the design of the HBS is the identification of the Primary Sampling Units (PSUs) as geographically defined non-overlapping groups or blocks of households based on the five yearly Census of Population (COP) Enumeration Areas (EAs) and their stratification into eight primary stratum groups. Therefore, it is correct to describe the HBS as a multi-stage stratified random area sample of private households. The are 34 geographical administrative areas which represent the 26 counties in Ireland The stratification procedure was undertaken in a number of steps or stages as illustrated by the table below. Stage 1 Sampling Unit Stratification
Selection of PSU = blocks Population density as based on COP Enumeration reported at the Census of Areas Population Selection of Household
2
The first stage of sampling involved, for each county, the identification of the COP EAs and their stratification in to eight population density stratum groups (i.e. nonoverlapping sub-populations). The number of the blocks, each containing approximately 75 households, in each EA is determined using COP data files. The blocks are sequentially numbered within each stratum. The population density stratum groups used in the 1999-2000 HBS were: Town Areas 1. County Boroughs 2. Suburbs of County Boroughs 3. Mixed urban/rural areas bordering on the suburbs of County Boroughs 4. Towns and their environs with populations of 5,000 or over (large urban) 5. Mixed urban/rural areas bordering on the environs of larger towns. 6. Towns and their environs with a population of 1,000 to 5,000 (other urban)
20
Country Areas 7. Mixed urban/rural areas 8. Rural areas In each county the PSUs or blocks are selected into the sample: taking into account the individual stratum distributions within a county. The number of PSUs or blocks to be selected were determined in the following three steps: Step 1 - This step computes the overall sampling interval in each county The number of blocks to be selected in each county is determined as a proportion of its share of the State’s population as recorded at the COP (i.e. a selection probability proportional to the county’s population). The overall county sampling interval can be defined as the inverse of the probability of selection of each person in a county, for a self-weighing design. Step 2: This step computes the overall sampling interval for the stratum groups in each county. The number of blocks to be selected for each stratum group within a county is determined as a proportion of their shares of the households in a county (i.e. a selection probability proportional to the stratum group's households within a county). The stratum group-sampling interval can be defined as the inverse of the probability of selection of each household in a stratum-group, for a selfweighing design. Step 3: Selection of the blocks in each stratum group in each county. In this step, a probability sample of the blocks to be surveyed within a stratum group in a county is selected by the method of simple random sampling without replacement. This procedure results in the selection of a sample of 2,549 blocks, which is approximately self-weighting
In the second stage of sampling, 4 households are selected for inclusion in the household sample by the method of simple random sampling without replacement in each block selected in first stage of sampling. Household substitution is allowed for non-response.
21
Appendix II
22
Table A: Detailed Expenditure and Measurement Values from a sub-sample of the HBS 1999/2000
Item Description Grams Litres Quantity Expenditure Expenditure No. of Conversion (£) (€) quotations Factor g/l/No. per €1
Food
BreadWhite Bread Soda Bread Brown Bread Other Bread Total Bread FlourWhite flour - plain White flour - self-raising Wholemeal flour Total Flour Biscuits Cakes & Buns Total Bread, Flour Biscuits & cakes Milk & Cream Fresh milk Fresh cream Total Fresh Milk & Cream 5,681.79 158.92 5,840.71 5,655.25 7,182.17 3,239 grams . Condensed milk Baby milk preparations Other milk (yoghurt) Total Solid Milk Products Total Milk & Cream Cheese 650 57,653 1,217,528 1,275,831 1,275,831 502,174 5,841 . 3.31 440.57 3,061.73 3,505.61 9,161 2,779.89 4.20 559.52 3,888.40 4,452.12 11,634 3,530.46 3 70 2,344 2,417 5,656 1,786 142.24 No. Eggs Butter, Fats & Cooking Oil Butter/Dairy Spreads Margarine Lard & Cooking Farts Total Butter & Fats 9,900 1,057.84 1,343.46 801 7.37 154.63 103.04 313.12 5,358.79 296.46 6,805.66 376.50 2,938 301 346,700 37,000 75,000 458,700 775,635 527,852 6,261,571 litres 0.83 0.42 13,539 17,194 10,760 181.56 19.65 249.41 450.62 3,733.25 2,974.53 230.58 24.96 316.75 572.29 4,741.23 3,777.65 182 23 36 241 3,073 1,871 1,503.59 1,482.64 236.78 3,095,370 302,762 617,159 484,093 4,499,384 3,681.60 388.16 592.11 1,718.41 6,380.28 4,675.63 492.96 751.98 2,182.38 8,102.96 2,915 380 618 1,662 5,575 grams 662.02 614.17 820.71 221.82
163.59 139.73
997,097 163,606 26,410 1,187,113
-
-
5,444.70 1,078.39 42.88
6,914.77 1,369.56 54.46 8,339
1,919 398 93 2,410
144.20 119.46 484.96
litres Other Fats & Cooking Oils Total Butter, Fats & Cooking Oil 1,187,113 403.35 403.35 623.80 7,189.77 792.23 9,131.01 336 2,746 0.51 6,566
23
Table A: Detailed Expenditure and Measurement Values from a sub-sample of the HBS 1999/2000 (contd)
Item Description Grams Litres Quantity Expenditure Expenditure No. of Conversion (£) (€) quotations Factor g/l/No. per €1 grams . 1,395.20 731.08 1,104.05 19.07 1,479.40 1,602.37 2,125.05 1,131.07 547.84 1,597.19 299.62 2,477.65 1,000.85 5,262.54 904.52 817.53 33.10 1,235.10 23,763 1,771.90 928.47 1,402.14 24.22 1,878.84 2,035.01 2,698.81 1,436.46 695.76 2,028.43 380.52 3,146.62 1,271.08 6,683.43 1,148.74 1,038.26 42.04 1,568.58 30,179 277 226 263 4 367 521 920 321 107 1,224 319 1,385 835 1,681 263 375 34 633 9,755 108.49 111.62 180.34 110.12 111.18 147.35 171.95 168.46 118.97 239.04 276.16 99.37 162.05 139.44 150.34 263.05 228.04 171.88
MeatSirloin Steak Round Steak Other Beef Mutton Lamb Pork Rashers Bacon - Uncooked Ham - Uncooked Sausages Pudding (black & white) Ham - Cooked Other Cooked Meat Chicken (cooked/uncooked) Other Poultry Minced Meat Liver Other Meat Total Meat Fish Fresh Fish Cod Haddock Plaice & Sole Whiting Other Fresh Fish Total Fresh Fish Frozen Fish & Fish Fingers Dried & Cured Fish Cod Haddock Other Dried & Cured Fish Total Dried & Cured Fish Tinned Fish Salmon Sardines Tuna Other Tinned Fish Total Tinned Fish Total Fish 33,348 9,551 55,467 11,269 109,635 459,681 1,579 7,443 6,544 15,566 192,228 103,638 252,869 2,667 208,896 299,860 464,050 241,992 82,775 484,884 105,083 312,695 205,976 931,915 172,703 273,115 9,586 269,610 4,614,541 . -
14,039 3,038 5,400 12,642 83,136 118,255 216,225
-
-
117.52 30.32 48.92 99.47 727.06 1,023 1,362.74
149.25 38.51 62.13 126.33 923.36 1,300 1,730.68
38 13 19 48 252 370 513
94.06 78.90 86.92 100.07 90.04
124.94
15.49 42.01 46.39 104
19.67 53.35 58.92 132
5 19 13 37
80.27 139.51 111.07
199.73 50.55 255.76 51.86 557.90 3,047.82
253.66 64.20 324.82 65.86 708.53 3,870.73
147 77 221 47 492.00 1,412.00
131.47 148.77 170.76 171.10
24
Table A: Detailed Expenditure and Measurement Values from a sub-sample of the HBS 1999/2000 (contd)
Item Description Grams Litres Quantity Expenditure Expenditure No. of Conversion (£) (€) quotations Factor g/l/No. per €1 grams
Vegetables Fresh Vegetables Potatoes Cabbage Tomatoes Cauliflower Brussels Sprouts Lettuce Carrots Onions Turnips Parsnips Red & Green Peppers Broccoli Other Fresh Vegetables Total Fresh Vegetables Dried Vegetables Tinned Vegetables Peas Beans Other Tinned Vegetables Total Tinned Vegetables Frozen Vegetables Peas Potatoes Other Frozen Vegetables Total Frozen Vegetables Total Vegetables 115,858 1,138,043 254,678 1,508,579 9,097,129 286,180 540,036 104,660 930,876 3,342,663 202,019 621,729 87,970 52,550 80,797 859,452 493,587 153,062 82,849 122,006 125,740 405,915 6,630,339 27,335 -
1,967.86 250.12 1,022.10 151.53 70.26 329.02 727.64 488.31 201.92 176.59 341.72 323.97 959.88 7,010.92 56.48
2,499.18 317.65 1,298.07 192.44 89.23 417.86 924.10 620.15 256.44 224.27 433.98 411.44 1,219.04 8,903.86 71.73
853 369 1,033 169 77 560 979 708 305 240 332 368 1,097 7,090 82
1,337.50 635.98 478.97 457.12 588.93 193.36 930.04 795.91 596.88 369.42 281.13 305.61 332.98
381.08
383.46 699.59 186.98 1,270.03
486.99 888.48 237.46 1,612.94
699 1,122 244 2,065
587.65 607.82 440.74
177.55 1,501.73 465.32 2,145 10,482
225.49 1,907.20 590.96 2,724 13,312
131 923 333 1,387 10,624
513.81 596.71 430.96
25
Table A: Detailed Expenditure and Measurement Values from a sub-sample of the HBS 1999/2000 (contd)
Item Description Grams Litres Quantity Expenditure Expenditure No. of Conversion (£) (€) quotations Factor g/l/No. per €1
Fruit Fresh Fruit – ApplesEating Cooking Oranges Bananas Grapefruit Lemons Plums Grapes Strawberries Kiwi Other Fresh Fruit Total Fresh Fruit Tinned & Bottled Fruit Pears Peaches Strawberries Other Tinned & Bottled Fruit Total Tinned & Bottled Fruit Dried Fruits & Nuts Raisins Sultanas Other Dried Fruits & Nuts Total Dried Fruits & Nuts Total Fruits & Nuts 26,487 19,602 83,601 129,690 3,498,272 52.98 32.82 334.72 420.52 5,588.20 67.28 41.68 425.09 534.06 7,097.01 58 50 293 401 4,943 393.66 470.28 196.67 91,595 27,109 12,357 107,421 238,482 155.14 46.55 28.71 198.10 428.50 197.03 59.12 36.46 251.59 544.20 165 63 23 255 506 464.88 458.55 338.90 426.97 grams
829,828 64,148 517,576 979,440 28,741 26,230 57,030 194,810 36,264
-
-
1,180.55 89.89 775.97 1,130.51 47.02 46.04 117.55 466.27 178.94
1,499.30 114.16 985.48 1,435.75 59.72 58.47 149.29 592.16 227.25
968 84 604 1,178 64 100 91 270 118
553.48 561.91 525.20 682.18 481.30 448.60 382.01 328.98 159.57
40,226
355,807 3,130,100
-
-
74.75
631.69 4,739.18
94.93
802.25 6,018.76
75
484 4,036
423.73
443.51 4,990
26
Table 4: Detailed Expenditure and Measurement Values from a sub-sample of the HBS 1999/2000 (contd)
Item Description Grams Litres Quantity Expenditure Expenditure No. of Conversion (£) (€) quotations Factor g/l/No. per €1 grams 1,709.35 974.38 131.74 1,270.22 624.12 173.07 118.75 3,728.23 448.68 10.38 644.32 298.14 148.64 102.46 57.72 31.23 51.47 2,423.24 4,294.72 1,742.34 2,557.49 122.15 1,522.41 360.39 370.06 211.43 761.21 628.19 1.70 25,518.23 2,170.87 1,237.46 167.31 1,613.18 792.63 219.80 150.81 4,734.85 569.82 13.18 818.28 378.64 188.77 130.12 73.30 39.66 65.37 3,077.51 5,454.29 2,212.77 3,248.01 155.13 1,933.46 457.70 469.98 268.52 966.74 797.80 2.16 32,408.14 916 380 79 866 538 97 122 1,738 407 16 791 180 191 97 81 40 46 1,847 2,722 1,257 1,670 59 693 363 294 179 733 428 2 16,832 litres 1,010.56 13,274.45 14,285.01 1,226.59 2,915.66 4,142.25 1,557.77 3,702.89 5,260.66 2,314 708 1,606 0.65 3.58 120.75 41.81 150.07 628.76 337.69 187.56 880.90 239.57 289.51 584.10 273.14 115.46 212.53 243.54 698.88 49.80 80.32 203.24 91.15 98.55 150.68 140.95 170.71 476.72 81.18 242.98 66.40 84.42 162.11
Miscellaneous Foods Tea Coffee Cocoa, Drinking Chocolate, etc. Sugar Jams & Marmalade Treacle & Honey Oatmeal Breakfast Cereals Rice Cornflour Other Cereals Prepared Baby Foods Jellies Custard & Blancmange Power Salt Pepper Mustard Sauces & Creams Sweets & Chocolate Potato Crisps Prepared Food Frozen Dinners Pizza Pasta Meat Cubes & Meat Extracts SoupTinned Packet Other Food Food Undefined Total Miscellaneous Foods Ice Cream & Juices Ice Cream & Ice Lollies Fruit & Vegetable Juices Total Ice Cream & Juices Soft Drinks Consumed at Home Consumed Out Total Soft Drinks Total Food 33,031,637 12,772.91 2.00 12,774.91 33,304 9,900 262,142 51,739 25,109 1,014,300 267,662 41,226 132,850 1,134,328 164,969 7,700 223,508 43,716 40,119 31,691 51,231 1,975 5,250 625,478 497,158 218,077 489,396 21,865 330,352 218,194 38,155 65,244 64,188 67,354 350 6,135,326 -
4,056.72 1.78 4,058.50 110,327
5,152.03 2.26 5,154.30 140,116
2,903 1 2,904 70,533
2.48 0.88
27