In selecting the items for the index, attempts were made to ensure that each sub-group contained all the most important items in that sub-group as well as a selection of the items of lesser importance to represent all other items covered by that sub-group. Consequently, many of the items in the index were chosen not because they are themselves particularly important but because they represent a wide range of consumer goods.
Major groups | Number of items (a) | Weights | Approximate number of price quotations |
---|---|---|---|
Food | 48 | 36.65 | 2970 |
Beverages and tobacco | 7 | 3.43 | 270 |
Clothing and footwear | 13 | 10.84 | 717 |
Rent, fuel, light, water | 16 | 13.07 | 239 |
Household items and operation | 24 | 13.70 | 1069 |
Medical care | 6 | 1.30 | 127 |
Transport and communications | 18 | 10.54 | 129 |
Other goods and services | 20 | 10.46 | 582 |
Total | 152 | 100.00 | 6103 |
Note: (a) Although national weights are not directly used in the computation of the index, they are given for reference purposes.
All types of shops are represented in the price collection exercise: large, medium and small. About 330 retail outlets are visited each month. Prices are collected by officers from Central Statistical Office in Gaborone who visit all the areas during the first two weeks of the month. About 6,100 price quotations are collected every month.
Domestic servants' wages are collected every three months in Gaborone. Some prices are ascertained through telephone calls. Discounts, sale prices and free-market prices for items which also have official prices are used whenever they exist.
Only cash purchases are taken into account.
Price collection is carried out in accordance with the Statistics Act of 1976. This Act requires that respondents provide the data asked for but at the same time requires Central Statistics Office to treat the data as strictly confidential and not to reveal it to any person, organisation or other Government department outside Central Statistics Office.
In this way, if a given shop has no stocks of the 250g pack of butter, no use is made of the price of the 500g pack. This is because it is not essential that a price is available for every item in every shop. Instead, all the price quotes for the 250g pack of butter from all shops within the same area are averaged and the average is compared with the base price for the item in that area. If a few shops do not have it in stock, this does not matter, provided at least one outlet in that area does have it in stock so as to provide an average price for it in that area. The following month the same item description for a 250g of butter appears on the questionnaire for that shop. This method assumes that the number of errors caused by pricing an item that does not exactly match the item description will be reduced.
The price relative for each item in each area is first calculated by dividing the average price for the current period by the average price for the base period. The relatives are then used to compute separate indices for low-, middle- and high-income group households in urban areas and for large and small villages in rural areas.
All-income group indices for the urban areas are calculated as the weighted averages of the three income group indices, taking into account the average consumption expenditure in urban areas. The national index is the weighted average of the urban and rural indices using the population of urban and rural areas as weights.
New series relating to urban areas, rural areas and the whole country (base November 1991 = 100) are now computed, but the relevant methodological information was not available in the ILO at the time the present volume was published.
Statistical Bulletin(Gaborone), (quarterly).
Idem: Cost of living index 1985
, November 1985.
Idem: Consumer Price Statistics
.