Watching the popular quiz programmes on Italian television one gets the impression that many Italians are still in love with the vecchia lira. The lucky winners are often reminded that their prize would run into millions of their old currency.

There could be historical, psychological and economic reasons for preferring the lira to the euro. On economic considerations, if, as many Italians assert, the prices of goods and services went up as a result of the introduction of the euro, then one could say that the average Italian was better off under the lira regime, in that he could buy more goods and services than with the same equivalent amount of euros.

The value of money is the power that money has in exchange for other commodities - its purchasing power. A general rise in prices or inflation, therefore, erodes the (real) value of a given amount of money.

Inflation is generally defined as the percentage rate of increase of the level of prices during a given period. Defining inflation as the rate of change in price levels, it can thus be measured by a price index that is designed to show the change in monetary value of a specified fixed basket of goods and services.

The Retail Price Index (RPI) is designed to measure the average change in the level of prices paid (inclusive of all indirect taxes) for consumer goods and services by private households. Malta's RPI should, more appropriately, be referred to as a Consumer Price Index (CPI), since both goods and services are included in the "basket".

It measures in index form, the monthly changes in the cost of purchasing a constant-representative basket of goods and services. It shows the current cost of this fixed "basket" as a percentage of the cost of the same identical basket at the base period. It is an aggregative weighted index of the Laspeyres type.

The RPI is based on a network of samples. The weighting basis of the index is itself derived from the average expenditure pattern of a national representative sample of private households. Sampling is also used in the periodic collection of prices since it would be impossible to price every brand or variety of every item available in all retail and service outlets in the country. In practice, it suffices to price sample varieties for a representative sample of items in a national representative sample of shops. As the index is intended to reflect price change only, the samples of identical varieties and shops are retained unchanged for as long as possible during the life span of the series.

It is of historical interest to read that, as early as 1925, the Department of Labour was already monitoring changes in the prices of a small basket of goods in order to ascertain whether the standard of living of workers' families had deteriorated in the aftermath of the 1914-1918 war. It seems, however, that the first scientific inquiry into the spending habits of 100 workers' households was held by the Commissioner of Labour and Emigration in 1936. A "cost of living" index was constructed in the following year. Other partial enquiries were conducted between 1943 and 1969. As from 1971, full scale or national surveys were carried out affording wider coverage of households drawn from different geographical localities. The last Household Budgetary Survey was held in 2000/1 and the results have been published by the National Statistics Office.

Changes in the spending patterns of households since 1980 can be read from the table above that has been derived from the results of the Household Budgetary Surveys carried out during the last two decades.

In the 1980/81 survey, it resulted that 41.91 per cent of household expenditure was devoted to food items. During the last survey, this proportion went down to 23.82 per cent. In contrast, in 1980/1, 9.93 per cent of total household expenditure was spent on transport and other items related to communications; this went up to 23.13 per cent in 2000/1. The last column includes the weights in respect of the groups of items that are being used for the compilation of the RPI as from January 2003.

How far can we equate the RPI with an index of inflation?

There are a number of qualifications in equating the RPI with an inflation index.

Firstly, in order to reduce inflation bias, chain indexes may be constructed in which the weights are changed frequently, possibly every year.

Secondly, the basket of goods and services should be large enough to be representative of a good cross-section of the population or of the various income groups.

Thirdly, since price indexes often include imputed prices, such as rents of owner-occupied housing, all such items have to be eliminated from the index and only items that are actually purchased should be included. Inflation occurs when prices of goods and services denominated in currency units are rising.

In the fourth place, measures of inflation need to be confined to items of goods and services that can be observed in different periods. Then again, an index which excludes non-market goods and services, or which gives them less weight (for example, when a service is supplied at a subsidised price by the government), would not measure the level of inflation.

Lastly, the price index used must be related solely to households' monetary transactions for the "direct satisfaction of individual needs and wants" and not to other units such as the government.

Having outlined some of the qualifications of the RPI as an inflation index, the reader might be interested to find out how the real value of the Maltese lira has changed (or better, decreased) over the past 22 years. This can be read in the table below.

In terms of purchasing power, the value of the lira, worth 100 cents in 1980, is now only 57c3. The biggest drop in its value occurred in 1980 and 1981 when the RPI registered increases of 15.8 per cent and 11.5 per cent respectively. In contrast, Malta, along with some other countries, had negative inflation rates in 1983-1985.

The RPI, as a measure of inflation, has been the subject of controversy for many years. Among the more recent investigations, one may refer to the Boskin Report to the USA Senate Finance Committee in 1996, wherein a number of recommendations were made to make the CPI a more acceptable measure of the purchasing power of money. Like the Boskin findings, I would not be surprised if this brief article will also provide an agenda for long discussions among pensioners and others with fixed incomes.

Mr Camilleri is chairman of the Malta Statistics Authority