Page 278 - DECO504_STATISTICAL_METHODS_IN_ECONOMICS_ENGLISH
P. 278
Statistical Methods in Economics
Notes • Price quotations should be obtained from the localities in which the class of people concerned
reside or from where they usually make their purchases. Some of the principles recommended
to be observed in the collection of retail price data required.
• Since prices form the most important component of cost of living indices, considerable attention
has to be paid to the methods of price collection and to the price collection personnel. Prices are
collected usually by special agents or through mailed questionnaire or in some cases through
published price lists. The greatest reliance can be placed on the price collection through special
agents as they visit the retail outlets and collect the prices from them. However, these agents
should be properly selected and trained and should be given a manual of instructions as well
as manual of specifications of items to be priced. Appropriate methods of price verification
should be followed such as ‘check pricing’ in which price quotations are verified by means of
duplicate prices obtained by different agents or ‘purchase checking’ in which actual purchases of
goods are made.
• In order to convert the prices into index numbers the prices or their relatives must be weighted.
The need for weighting arises because relative importance of various items for different classes
of people is not the same. For this reason, the cost of living index is always a weighted index.
While conducting the family budget enquiry the amount spent on each commodity by an average
family is ascertained and these constitute the weights. Percentage expenditures on the different
items constitute the individual weights’ allocated to the corresponding price relative and the
percentage expenditure on the five groups constitute the ‘group weight’.
• The Sixth International Conference of Labour Statisticians recommended that the pattern of
consumption should be examined and the weights adjusted, if necessary, at intervals of not
more than ten years to correspond changes in the consumption pattern. The index also does not
take into account changes in qualities. Unlike changes in consumption pattern changes in
qualities of goods and services are more frequent and when a marked change in the quality of
items occurs appropriate adjustment should be made to ensure that the index takes into account
changes is qualities also. But in practice it is a difficult proposition to follow and, therefore,
constant qualities are assumed at two different dates which again is a shaky assumption.
• The consumption pattern derived from the expenditure data of a sample of households covered
in the course of family budget enquiry has to be representative of all the items in the average
budget, the localities from which price data are collected have to be representative of all the
localities from which the population group make purchases, the retail outlets from which prices
are collected have to be representative of all the retail outlets patronised by the population
group, etc. However, it is often difficult to ensure perfect representativeness and in the absence
of this the index may fail to provide the real picture.
• While taking the sample random sampling is seldom used. This is so because to sample from a
population of literally millions of commodities and services, the random procedure could neither
be practical nor representative. Typically, indices are constructed from samples deliberately
selected. This is likely to introduce errors and every effort must be made to minimise these
errors.
• A large number of methods have been designed for constructing index numbers and different
methods of computation give different results. Very often the selection of an appropriate formula
creates problems and in the interest of comparability, it is necessary to ensure that the same
formula is adopted over a period of time for constructing a particular index. There is no index
number method which is most satisfactory from all the various points of view which may
logically or practically be taken. Index numbers are averages, and all averages are basically
compromises between opposing extremes or forces.
21.4 Key-Words
1. Combinations : The number of ways objects can be selected without regard to order.
2. Combinatorics : The branch of mathematics dealing with the number of different ways objects
can be selected or arranged.
272 LOVELY PROFESSIONAL UNIVERSITY