Reg Rumney looks at the latest row about how official inflation is measured
Once again the official inflation figure supplied by the government’s Central Statistical Services has been called into question.
Cosatu said this week a study by the union-aligned research group Naledi confirmed its suspicions that the official inflation figure is inaccurate.
The CSS is adamant that its inflation figure, measured in terms of the Consumer Price Index, is accurate. The year-on-year inflation rate, as measured by the Consumer Price Index, was a whisker under 10 percent in December, giving an average inflation rate for the year of nine percent.
The Naledi Research Report, by economists Neva Seidman- Makgetla and Rob Rees, says: “In 1994, the official consumer inflation rate, as measured by the CPI, ran at nine percent. Union members argued, however, that in their experience these figures were too low. In the event, our research finds that in 1994, the official inflation rate understated the actual increase in prices for the low-income group by up to 3,5 percent a year. Moreover, the official figures underestimated inflation for at least four of the past five years.”
Naledi’s argument is that the CSS has underestimated the share of food in spending by poor households. Food inflation has been far higher than for other goods in recent years.
Low-income groups have a low weighting in the CPI, despite 80 percent of households falling into the low- income category. Hence, says Naledi, the official inflation rate is inaccurate.
Not so, replies CSS head Treurnicht du Toit. He contends the CSS publishes enough information in enough detail for anyone to calculate the particular inflation rate for their own household. “This is no reason to tell us we are wrong,” he says.
In short, the CSS does not maintain the basket of goods and services in its Consumer Price Index is representative of each group or individual.
There are two sides to the measurement of inflation, he says. The one is the measurement of the price changes of actual goods and services, such as shoes and haircuts. Du Toit is confident these are accurate. The other side is the weighting, which he says is designed to reflect the average spending patterns of households. The weighting is decided every five years according to a survey of the spending habits of the population.
The current weightings are derived from a 1991 Human Sciences Research Council survey, says Du Toit. The next survey, to be done by the CSS in October, will coincide with a survey to extract statistics on unemployment and the informal sector for the RDP.