The start of the winter has been awash with reports about wage bargaining. Central to the discussions is the change in the cost of living, reflected by the consumer price index (CPI), which includes the goods and services that a household pays for over a year.
It does not refer only to necessities, but also includes items bought infrequently, such as furniture and vehicles.
This week Statistics SA released the figures for May, which showed that the increase was 5.7%.
Wage increases should compensate employees for higher prices so that they are able to maintain their living standard. Although many bargaining council agreements stipulate that the index should be a key consideration for wage increases, there are often complaints during negotiations that the official consumer price inflation is much lower than the actual inflation experienced by the average worker. Frequently, these are backed up by reports of visits to the local supermarket by some well-meaning research institution.
Of course, there are other factors, such as overall gross domestic product growth and the health of the sector or a specific firm that are also considered.
Standard of living
Generally, employees have not done badly in maintaining their standard of living. The earnings of employees in the formal sector increased by between 5% and 9% over the past year, according to figures from Stats SA’s Quarterly Employment Statistics survey. It compares favourably with the 4% to 6% inflation rate over the same period.
One of the objectives of the South African CPI is to measure changes in the cost of living. It has been done in most countries since the late 19th century. The international standards for CPIs are regulated by the International Labour Organisation and the South African index fully subscribes to the organisation’s resolutions on how an index should be compiled.
The cost of living of a household is determined by the spending pattern of all households in a given period. In South Africa, it is captured by a household income and expenditure survey, which scientifically surveys a large representative cross-section of families for a year. This forms the basis of the list of products and services (basket) and their relative importance (weights) that comprise the monthly CPI survey.
The CPI fixes its basket and weights for a specific time. Normally, these are refreshed every five years. The current weights are based on the 2005-2006 income and expenditure survey and were first used in the CPI in 2009. In 2013, a new basket and weights will be introduced, based on the 2010-2011 income and expenditure survey. Its results will give a clear indication of how households have changed what comprises their cost of living.
Two main factors drive changes in spending patterns. The first is changes in real income (that is, over and above the rate of inflation). Increases in income raise the proportion of spending on services and recreational and luxury products, known as the income effect.
The second key shift is driven by the differences in price increases between different goods. So, if the price of beef increases more than the price of chicken, consumers will shift expenditure to chicken, known as the substitution effect.
If we assume that income gains have been modest, it means the substitution effect would dominate. The effect of a fixed basket means that the CPI is likely to overestimate the real cost of living because it assumes that, with higher inflation, households will still buy the goods.
Counterbalancing this is that the prices of some products for which there are no obvious substitutes, such as petrol or electricity, have increased significantly in recent times. We do not really know the extent to which public transport or the use of alternative energy sources may have reduced consumption of these products.
It is in tracking these substitution effects that the CPI does reflect some limitations as a cost-of-living indicator. A true indicator would reflect changing spending patterns each month and the changes in prices. Unfortunately, this is logistically impossible.
However, the CPI uses a “geometric mean” formula to average the price changes at the most basic level of the index, for example, to get the average price change of brown bread in Polokwane. A geometric mean allows for a modest amount of consumer substitution as relative prices within item categories change, which strengthens its role as a cost-of-living indicator.
Quality of government services
Theoretically, a complete cost-of-living measure would also monitor other factors that affect households’ standard of living, such as the quality of government services and the environment, but in practice it is well-nigh impossible.
The other way to improve the relevance of the CPI is to update the weights more often. A continuous expenditure survey is run in several countries. It allows the CPI weights to be adjusted on an annual (in the United Kingdom) or biennial (in the United States) basis. These surveys are expensive, but large amounts of money ride on the inflation rate.
The primary element of the CPI is the monthly price survey in which more than 65000 prices of about 400 products are recorded in 25 different towns and cities in South Africa. It ensures that the prices paid really are those that go into the calculation. For example, in May, Stats SA recorded the price of 1206 loaves of bread, 249 bags of maize meal, 317 bars of soap, 335 televisions and 471 items of women’s underwear (the sample size is determined by both weight and price variability).
The CPI reflects the spending patterns of South Africans as a whole and is unlikely to reflect any individual household’s cost of living exactly. But if you are interested in calculating the likely CPI for your household, use Stats SA’s personal inflation calculator on its website.
Patrick Kelly is executive manager of price and employment statistics at Statistics South Africa