IRISH CASE STUDY:
Chapter 2: The Tools of Economic Analysis & Chapter 28: Inflation

The Consumer Price Index: Not a True Cost of Living Index
by Bernadette Power, Department of Economics, University College Cork.

This case focuses on sources of error in the consumer price index (CPI), the current measure of inflation in Ireland. It provides a link to the some issues and concepts discussed in chapter 2 by examining the ability of facts and data collected to explain economic relationships. It also deals with the construction of index numbers, the concept of purchasing power and indicates how measurements of economic phenomena can be oversimplified. Students should question whether or not we should not use these measurements to test economic relationships and realise how difficult it is to improve them.

The rising rate of inflation in Ireland, measured by the Consumer Price Index (CPI) has raised much controversy regarding the Programme for Prosperity and Fairness (PPF) and the indexing of Public Sector wages for changes in the cost of living. The PPF was predicated on certain growth rates being achieved (5.5% p.a.), agreed tax reductions being implemented over the period of the agreement, coupled with inflation being maintained at a given level (approx. 2.3% p.a.) and other non pay elements. Growth has exceeded the conservative levels outlined in the PPF and projections for this year and next year were 8.5% and 7% respectively. In addition inflation has risen far higher than expectations over the previous months at twice the level indicated in the PPF.

The CPI is supposed to represent the movement in prices of a basket of goods and services that the average Irish family buys. A rise in the CPI measure of inflation by more than increases in take home pay over a period of time means consumers are less well off. However, it is far from a perfect measure of changes in the cost of living as it overstates the true level of price change under certain circumstances and understates it under other circumstances. A true cost of living index measures how much extra money an individual requires in order to achieve the same level of satisfaction from their purchases if prices rise from one period to the next. However, the CPI is not constructed with regard to a base level of satisfaction. This leads to errors in the measurement of changes in the cost of living. In fact the CPI is designed such that it fails to take into account the manner in which households change their pattern of expenditure in response to changes in price, incomes, family composition, tastes or market composition.

Some of the sources of error in the CPI are as follows;

Outlet Bias
The CPI index is based on prices recorded at a wide range of outlets, both expensive and cheap. However, it ignores the tendency to move purchases from expensive to cheaper outlets such as from local shops to large suburban supermarkets. Such cheaper forms of shopping is not reflected in the CPI. Factors that can also reduce the list price or cost of goods to consumers are manufacture issued coupons or value cards issued by large retailers such as Tesco and Dunnes Stores and should also be reflected in the index. Other factors such as search costs, the costs of writing contracts and delivery costs increase the list price of goods.

New Good Bias
The CPI index is late in adding new goods to the basket. Therefore there is a significant time lag between the introduction of a product onto the market and its incorporation into the index. The consumer is better off as there are a greater variety of goods on offer, however the typical price index does not reflect this. The prices of many new goods generally fall rapidly after the initial introduction period and therefore the price index could be overstated in this period if these goods are included in the index.

Quality Bias
The average quality of goods tend to increase over time but little has been done to include quality improvements in the CPI measure causing it to overstate the true rate of increase in the cost of living. Today's £1,200 computer is far superior to a £1,200 computer in 1990. Furthermore there is a difficulty in distinguishing changes in the quality of a good and new products at times. For example, was the CPU a new product or a quality improvement on the mainframe?

Substitution Bias
As the CPI index measures changes in the cost of buying a given basket of goods purchased in the base year and not a base level of satisfaction it does not allow for the possibility of substitution away from those goods whose relative price has risen to cheaper substitute products. This source of error may be small in countries which frequently update the weights on which the index is based such as in the United Kingdom where they annually update the weights. In Ireland the weights are updated less frequently so substitution bias overstates the cost of living of consumers. However, if it is fair to adjust for quality gains, quality loss should also be included. Chicken, though a substitute for steak, is not steak.

Quality of Life
The CPI does not capture quality of life reductions that entail new costs, e.g. the cost of an alarm system necessitated by deteriorating social conditions.

Changes in Tastes
The weights for the CPI are based on the consumption patterns of a representative consumer. However price changes effect different types of households differently by virtue of differences in their expenditure patterns. Individuals are not affected by increases in the price of cigarettes if they don't smoke. Hence, there should be a separate index for non-smokers. The CSO produce a CPI index that excludes tobacco costs but other preferences e.g. vegetarians and teetotallers are not accounted for.

One Period Analysis
The CPI is based on a one period analysis of prices and is therefore essentially based on a static theory. The correct treatment of durable goods within such an index is as a flow of services. This approach is adopted for housing costs in the current CPI index where the cost of the flow of housing services consumed over the measurement interval is included using a rented equivalence approach (rented income is included not the value of the house). This approach is not employed for all durable goods that yield a flow of services such as motor vehicles, televisions and refrigerators. Therefore not all-durable goods are treated similarly. The measurement of housing cost is not perfect also as the rising cost of mortgages is based on mortgages on houses, which were in existence in 1996. House prices and the size of mortgages have increased in Ireland considerably since then. Furthermore increases in rental prices lag increases in house prices, as landlords cannot continuously increase rents to reflect changes in the market value of their rental property but renegotiate at intervals. Also with high house ownership in Ireland the size of the rental market is smaller than that in other countries thus the rental equivalence approach which accounts for changes in the cost of housing may not truly reflect increases in house prices as they occur.

Further errors also exist. Items such as income tax and social insurance contributions have an important impact on household budgets but are not included in the CPI. The CPI includes indirect taxation which is extraordinary. If the finance minister taxes us indirectly in the budget through raising the prices of cigarettes, the CPI rises: if he raises income tax in the budget it does not. Indeed the government recognises this irrationality by publishing an index which excludes indirect taxation. Similarly, interest rates are treated unusually. Interest rates are used as a mechanism for controlling inflation so but when the European Central Bank raises interest rates this in turn leads to rises in mortgage interest rates which increases the CPI index. Again an index that excludes mortgage interest is produced to oversee the extent of this problem. Conversely the CPI excludes important elements of our national personal expenditure just because they happen to be run by the government - the General Medical Scheme (GMS), education and roads. Even though the government undertakes these expenditures it does not mean that it has no effect on inflation as taxes will rise to reflect the level of inflation generated by government run services.

The Household Budget Survey (HBS) on which the weights in the CPI are based are widely acknowledged as unreliable. The weights were last updated in November 1996 based on the HBS conducted in 1994/1995. Spending patterns may have changed since 1996. The CSO admits that the weights also have to be fudged for drinks, sweets and tobacco as they are underreported in the HBS but it may well contain other errors also.

Thus, CPI inflation data do not translate into a cost of living index. It simply measures increases in consumer prices. These errors could mean a one or two percentage point difference in the measure of inflation. This explains why many people feel their pay packets are not stretching as far as they used to. The problem is that no such index of price changes exists at present. However there can be little doubt that such an index could prove useful for wage negotiation between the social partners. The wages increases under the PPF were based on the inflation running at around 2.3% percent. Most taxpayers would have assumed that inflation means cost of living but if that is significantly higher many could feel cheated. If wages are to be indexed linked we need to get a correct measure of changes in the cost of living. Otherwise it will lead to dissatisfaction among employees and potential strike action as we have seen by bus drivers, nurses, gardai and many more groups. Furthermore if governments are to use the CPI as a crucial policy variable in examining the inflationary impact of government policies we needs to get right also.

Consumer price Commodity Group Indices, July 2000

 

 


References

Consumer Price Index, July 2000, Central Statistics Office, Ireland.

Pay Policy Review is Inevitable Given Projected Level of Inflation, Irish Examiner Monday 17/7/2000.

Programme for Prosperity and Fairness, Dept of Government, Ireland.

QUESTIONS FOR DISCUSSION

1. What is a true cost of living index supposed to measure? Why is the Consumer Price Index not a true measure of changes in the cost of living?

2. What is the effect of excluding an item from the consumer price index? How important are the weights in the calculation of the index? Examine the Table below in answering this question.

3. Discuss the importance of developing a true measure of the cost of living and some of the difficulties involved in undertaking this task.