Implications of the new Consumer Price index

Picture shows a busy market place. There is concern that the new index is not representative enough of the impacts of inflation on the poorest groups in society – those groups which usually suffer most from increases in the cost of living, particularly on food items

The Colombo Consumer Price Index (CCPI) is the most widely used tool for measuring changes in the cost of living or inflation which are important in formulating economic policies and makinginvestment decisions, so the Department of Census and Statistics’ revised Colombo Consumer Price Index (CCPI(N)) marks a significant milestone.

The CCPI has been considered the ‘official CPI’ since 1953 and has been used as the official indicator of changes in cost of living allowances payable to government servants and workers in other industries under the operation of the Wages Board Ordinance. The weighting pattern used in the index was based on the average expenditure of a sample of 455 working class households, as identified in the Colombo Family Budget Survey 1949/50, and revalued at 1952 prices. Economic analysts and trade unions have long argued that a new CPI is needed to eliminate many inherent problems and limitations, such as representation of population, geographical coverage, the basket of items and the weights of those items. The new index introduced by the Department of Census & Statistics (DCS), CCPI(N), is based on 2002 Household Income and Expenditure Survey (HIES) data and represents more up to date consumer patterns for a much larger sample size, as well as an increased coverage area within Colombo for price collection.

Whilst there was overall agreement that the DCS’ new index was welcome, the recent debate at CEPA’s 33rd Open Forum (to discuss implications of the updated index with the panellists including D. C. A. Gunewardena, Director Price Statistics from the DCS, Dr Harsha De Silva, Lead Economist from LIRNEasia and Deshal De Mel, Research Officer from the Institute of Policy Studies) highlighted some issues which continue to concern analysts and users of the data. The two main areas of concern were which group the index is representative of, and how the index is used.The new index is no longer based on a ‘working class’ sample of households, and reflects a conscious decision on the part of the DCS to collect data from all socio-economic groups, creating an overall average basket of goods and average weightings.

There is, therefore, concern that the index is not representative enough of the impacts of inflation on the poorest groups in society – those groups which usually suffer most from increases in the cost of living, particularly on food items. Including the most affluent households in such an average is known to skew the data towards the higher end of the scale, particularly the weightings, since the poor spend a higher proportion of their income on food, for example, than the better off. This means that an index which has a lower weighting on food and a higher weighting on non-consumables, such as clothes, communication, education and transport, may not give an accurate picture of how poorer households are experiencing changes in prices. The index also continues to be limited to Colombo urban consumer patterns and prices which makes it difficult to judge the impacts of inflation on rural and estate populations in the rest of the country (who make up the majority of the population, and the majority of the poor).

This is highlighted in the difference between the proportion of income spent on food in the Western Province (29.7%) and in Sabaragamuwa Province (45.6%).

In many countries this problem is addressed by using purpose-defined indices which reflect the reality of consumption patterns for a specific group, e.g. the USA has two major indices, one which covers all urban consumers (87% of the population) and the other which covers only urban wage earners (only 32% of the population) and is used to determine increases in cost of living allowances and social security payments, since this is the group most vulnerable to those payments. So, should Sri Lanka also have such purpose-defined indices? Two issues would need to be considered; one is the problem of data collection – do we have the systems in place to be able to collect data for more than one index and from areas outside Colombo? The DCS has already acknowledged that it has difficulty collecting market prices outside the capital, hence the reliance upon the Colombo CPI rather than the Sri Lanka Consumers’ Price Index (SLCPI), introduced in 1997 but never been widely used. The second issue is that of usage, how is the index currently used and is there a demand for purposive indices?

Whilst the CCPI is widely used as an indicator of how the economy as a whole is faring, the extent to which it is used in wage bargaining appears to be less influential.

Cost of living allowances payable to government servants and industries which fall under the Wages Board Ordinance are linked to the CCPI, but in most other private and informal sectors the market and trade union-led collective wage bargaining agreements tend to drive wage rates. This means that a significant proportion of people’s wages are not directly impacted by the CCPI, and pensions are also not officially linked to the index. Further, without legislation to approve the change, the revised CCPI(N) will not be used to calculate changes in cost of living allowances, the value of which are still linked to the old ‘official’ index. The financial implications of revaluing the index point’s worth, or encouraging wider usage of the index in wage agreements, are onerous in today’s high inflationary environment and therefore unlikely to occur.

This highlights the reality of the situation; whether we are using the old CCPI, the new CCPI, the SLCPI or hypothetical purposive indices for different groups – the overall picture of recent inflation will be the same, rapidly increasing. For this reason some may choose to either discredit or welcome but not act on the new index since such wage increases are considered unsustainable. Arguing over the detail of differences between socio-economic groups may, therefore, be considered a red herring to distract from the overall trend. It is well known that when new or revised indices are introduced in any country, the initial tendency is for them to portray a lower figure than the old index because the new index reflects recent substitution behaviour by consumers (replacing more expensive goods with cheaper alternatives) in order to cut down on the cost of living. This was the case with the retrospective calculations of the CCPI(N), and may, in part, explain the extent to which Central Bank Governor Nivard Cabraal has welcomed the new index, asserting in a recent interview on ETV’s LBR that “inflation is not as high as it has been recorded in the past because the [old] index we used was quite faulty”. However, the honeymoon period may not last as long as hoped, with the CCPI(N) December figures already showing a higher rate than the old CCPI as the CCPI(N) has more weight on petrol, gas and other import driven sensitive items. Recent price increases are likely to drive this up even further.

The debate also noted that despite ILO recommendations dating from 2003 that consumer price indices should cover all types of consumer goods and services of significance to the reference population – “without any omission of goods that may not be legally available or may be considered socially undesirable”, tobacco and alcoholic beverages are excluded in the CCPI(N)’s basket of item. The prices of these items are frequently revised therefore it could have a significant impact on inflation figures, but in an attempt to de-link any cost of living increases to the consumer’s ability to purchase alcohol and tobacco the government specifically requested that they be left out of the index.

Perhaps in a lower inflationary environment, discussion on the need for and use of indices which better reflect the consumption patterns of different socio-economic groups may be welcomed and taken more seriously by both private and public sector employers. The unwillingness of many to act on the figures reflected in the new index appears to be based on the simple fact that they don’t like the emerging picture rather than because of criticisms of the index’s technical faults.

However, it is vital to remember that inflation hurts the poor first and foremost, and for this reason the shortcomings of the new index in reflecting the impacts on the poor should not be taken lightly. - (Contributed by CEPA)