Jasim Uddin Haroon
The inflation data measured on the consumer price indices (CPI) sparks frequent debate that the official measure either understates price increases or exaggerates them.
Many argue that product-basket in Bangladesh is small resulting in inaccurate rate of inflation. On the other hand, some argue that base year is too old to reflect the real situation of price changes in the market.
Again many alleged that the field workers of the Bangladesh Bureau of Statistics (BBS) do not visit the markets leading to 'distorted' rate of inflation.
The objective behind the writing this piece of article is to answer questions relating to the CPI, its construction, methods and other issues concerned.
According to Economics literature, all of the changes in market prices that have occurred during the current year or current month due to inflation or deflation is measured under a base year that generally chosen a normal year free from economic turmoil.
While calculating CPI which measures changes in the price level of consumer goods and services purchased by households uses a base year. The BBS, the national statistical organisation (NSO) had been using 1995-96 as base year. This base was 17 years old. So the allegation of old base year is true. Products and services change fast. Uses of mobile phones, internet, drinking water have become important items in civic life in recent years. So revisit of products and changing of base year were of paramount importance.
The NSO has introduced a new base year --- 2005-06 since July of this year. But it has been producing CPI and inflation data by using old base year as well.
Inflation which is defined as a rise in the overall price level, and deflation is defined as a fall in the overall price level measured under a base year represent much accurate picture if its base year becomes near the calculating period.
The base year is always equal to 100 as per the rules of statistics. The 100 were divided into the products and services. But the distribution of weights is very important. If we give low weight on important items will yield distorted CPI and inflation.
In Bangladesh rice has the highest weight in food basket and rise in the prices of rice influences the food index.
BBS has upgraded its basket with the new base year. Now rural basket consists of 318 items against 182 in old base line. The new urban basket consists of 422 products against 271 products in base year 1995-96 base year.
Practice in Bangladesh
The BBS compiles CPIs on monthly basis. It takes market data from its selected market both in urban and rural areas upto 21st day of each month.
The monthly prices of different products are used to compute CPIs. BBS computes three types of CPIs --- national, rural and urban areas. To compute the three CPIs, all goods and services are classified into eight groups. The groups are: 1) food, beverage and tobacco 2) clothing and footwear 3) gross rent, fuel and lighting 4) furniture, household equipment 5) medical care and health expenditure 6) transport and communications 7) recreation, entertainment, education and cultural services and 8) miscellaneous goods and services.
This is rather a technical issue and this the crucial job for new base year as the commodities included in the index are not all of equal importance. BBS like al other national statistical organisations puts weights of the different products in accordance with their importance.
In Bangladesh, weighting patters of CPI national: Base year 2005-06=100. i.e. National =100, Rural= 64.89, Urban= 35.11.
There are a large number of formulae used in calculating CPIs. In statistics, Fischer's formula is an ideal. But, many NSOs uses Laspeyres formula as it emphasises on the prices. BBS uses Laspeyres formula. The formula is:
= sum of, Pn= price in the current year, Po= price in the base year, Wi= weight at the ith item and W= weight of the group.
Coverage and source of data
The price data are collected from the selected markets and outlets by the trained field staff of Bangladesh Bureau of Statistics.
(i) Market: For price data collection 140 markets has been selected all over the country. The selected number of urban markets are (64+12)=76 and the number of rural markets are 64. The field staff of Bangladesh Bureau of Statistics collects the price data from 64 urban and 64 rural markets (One urban and one rural market from each district) all over the country. The field staff of prices and wages section of National Accounting Wing are collects the price data from selected 12 markets in different parts of Dhaka metropolitan city.
(ii) Outlet: For each item price is collected from three selected outlets. There are some big outlets in a market which have varieties of consumer items, generally that type of outlets is selected to collect the price data.
(iii) Collection Procedure: Prices are collected during the peak hours of transaction. During the process of data collection, the price collectors are supposed to act as a true buyer, to act and bargain like a true buyer and to make an actual purchase. Price collectors should make sure that they are pricing the correct product based on the product specifications in the CPI basket. Price data are properly scrutinized in immediately after collection.
1. National inflation on October in 2012 (on point-to-point basis)
= 181.26÷ 171.23x 100 = 5.86 (Monthly)
Here, 181.26 is the national index of October in 2012 and 171.23 is the national index of October in 2011.
2. Again we can calculate monthly inflation as follows:
Monthly inflation for October in 2012
=181.26÷179.44X 100=1.01 per cent
Here 181.26 is the national index of October in 2012 and 179.44 is its previous month's (September 2012) index.
3. We can calculate yearly inflation as follows:
National inflation of 2011-12= 170.19 ÷156.59 x 100 =8.69 per cent
Here, 170.19 is the index of 2011-12 and 156.59 is the index of 2010-11.
Currently the BBS produces CPIs and inflations using both old and new base years. The NSO will use only new base year --- 2005-06 from January. i.e., in the first week of February, BBS will provide January inflation of 2013 under the new base year.
In fine, my observation is that the overall function in calculating CPIs and inflation is almost in line with international practices. But, my serious reservation is that on the types of inflation. The BBS calculates two types of CPIs and inflations, namely food and non-food. This is a very obsolete practice. Advanced nations and even developing nations are preparing core inflation and headline inflation. Headline inflation is more volatile than core inflation. Many consider non-food inflation as core inflation. But in theory, it is not correct as it (non-food) does not include food prices.
Core inflation represents the long run trend in the price level and it is very important for the policy markers. In measuring long run inflation, transitory price changes should be excluded. i.e core excludes items frequently subject to volatile prices, like food and energy.
Headline inflation is a measure of the total inflation within an economy and is affected by areas of the market which may experience sudden inflationary spikes such as food or energy prices.
The authorities concerned might consider introduction of these two types of inflations by avoiding food and non-food inflation.
The writer is special correspondent of the FE