Why are retailers' food inflation numbers much lower than official figures from Stats SA?

Why are Statistics SA's official food price inflation numbers so much higher than those listed by popular supermarket chains?

Just two weeks ago, Shoprite reported their internal selling price inflation at an average 0.7% for the second half of 2025.

Releasing the February consumer price inflation figures this week, Stats SA said food inflation was at 3,7%, falling from 4,4% in January.

RELATED: Meat price decrease helps cool food inflation, but looming fuel hikes could change consumer outlook

In an opinion piece published by Business Day, the executive chairperson of consulting firm Krutham explores what gives rise to this discrepancy.

Stuart Theobald also cites Pick ’n Pay’s numbers - its inflation measure for the six months to end-August 2025 was 2.1%, against Stats SA’s 4.6% for the equivalent period.

"Together, these two chains dominate food retail in South Africa. How is such a discrepancy possible?"

In conversation with Stephen Grootes, Theobald says that, substantially, it comes down to a difference of methodology.

As he explains, what Stats SA does involves measuring a basket of goods containing 65 different items for food and beverages. The data is collected by field workers sent into stores to look at prices and record them.

While that basket is meant to represent the typical consumer, it is determined only every five years, he emphasizes.

"The last time was back in 2023 and it's based on a survey of households who tell Stats SA what they consumed. But things change: In 2023 we were in the middle of loadshedding, there were likely more candles in that basket... and of course prices shift in the interim. People might consume more of things that get cheaper and less of things that get more expensive."

And, according to the standard approach of measuring prices every month via a basket determined four years ago, you're likely to become detached from what consumers are actually doing day-to-day in the stores, Theobald adds.

Added to this, stores are quite dynamic in their pricing, he says.

"The actual prices that consumers are paying, as Shoprite will know from its systems, can diverge considerably from the prices that Stats SA is measuring. And that's what seems to have happened in this example."

While Theobald acknowledges that it can be very difficult to measure what people are paying in the informal sector, he points out that this is not the case in the formal sector.

Could Stats SA be using the data available from retailers to make their own numbers more reflective of reality?

It is an idea worth exploring, he says, highlighting that this kind of exercise is already being done in other parts of the world.

"For example, there is a project in the UK between the Office for National Statistics and supermarkets to see if they can set up channels to directly obtain scanning data from the tills in those stores as probably a better and potentially cheaper way of collecting data than sending out field workers."

"In today's communications environment there are potentially straightforward ways that Stats SA could obtain data from the likes of Shoprite, Pick n Pay and Spar who between them account for way more than half of all the beverages and food consumed in the country... So I think the answer should be yes - there are things that could be done and we should be exploring them."

Scroll up to the audio player to listen to Theobald's argument and click here to read his opinion piece