Administered-price inflation accelerated in September, but the measure excluding fuel and paraffin edged lower. That distinction changes what businesses and policymakers can reasonably conclude from the 13% annual figure.
Key findings
- Administered-price inflation accelerated while inflation excluding those prices stayed unchanged.
- Excluding fuel and paraffin changes the direction of the latest movement, but a gap remains.
- SBL analysis: the latest acceleration does not establish an across-the-board tariff surge, nor does it show that non-fuel pressure has disappeared.
The evidence: two different price paths
Stats SA’s September 2026 CPI release, published on 21 October, puts annual headline inflation at 4.5%. Administered-price inflation reached 13.0%, compared with 3.3% for the basket excluding administered prices. The comparison below uses the agency’s September and August releases.
| Measure | August 2026 (%) | September 2026 (%) | Change (percentage points) |
|---|---|---|---|
| Headline CPI | 4.4 | 4.5 | +0.1 |
| Administered prices | 10.6 | 13.0 | +2.4 |
| CPI excluding administered prices | 3.3 | 3.3 | 0.0 |
| Administered prices excluding fuel and paraffin | 7.1 | 7.0 | −0.1 |
Source: Stats SA P0141, Table A, August 2026 and September 2026. Rates compare each month with the same month a year earlier. Changes are SBL subtraction, not monthly price increases. Figures are rounded as published; not all prices are surveyed monthly.
Administered prices include government-set or significantly government-influenced prices, including some charged by private suppliers. Stats SA’s release glossary lists fuel, electricity, municipal services and other items. This is not a category for all essentials, and classification does not establish what caused an increase.
Why fuel matters, and why it is not the whole explanation
Stats SA’s 21 October explanation reports fuel inflation of 27.9% over the year and a 6.2% monthly increase. Passenger-transport inflation reached 12.3%, up from 11.6% in August. Food and non-alcoholic beverage inflation, by contrast, was 1.2%.
The Department of Mineral and Petroleum Resources’ 31 August statement, covering adjustments effective 2 September, identified international oil and product prices, the exchange rate and domestic pricing components. It reported that rand appreciation partly offset higher international product costs. This is evidence about September’s mechanism, not a claim about today’s pump price.
SBL analysis: separate the acceleration from the level
The administered-price gap against the rest of the basket widened from 7.3 to 9.7 percentage points. Meanwhile, the administered measure excluding fuel and paraffin eased slightly. These are different signals, and neither should be discarded.
First, the overall acceleration cannot fairly be presented as evidence that every regulated bill suddenly rose faster. Removing the two fuel categories changes the direction of movement. Second, slower inflation in the remaining administered basket does not mean its prices fell over the year. Its annual rate remained above headline inflation.
This is a composition test, not a complete causal decomposition. It helps identify which claim the data can support. It cannot tell us how much of any individual tariff increase was needed to cover maintenance, wages, investment or inefficiency.
Business and public-finance impact
The policy question is consequently more precise than whether administered prices are “too high”. International fuel shocks and local tariff decisions have different causes and different possible responses. Treating them as one problem risks choosing a remedy that does not address the relevant cost.
For example, reducing a domestic charge would not remove imported product costs. Conversely, an improvement in international fuel prices would not automatically reverse a separately approved local tariff. These are economic mechanisms, not forecasts or proposals announced in this release.
Tariff restraint can benefit paying customers. But if a charge supports a public service, reducing it without a credible funding alternative could shift costs elsewhere or put service delivery at risk. Higher tariffs can also worsen affordability and collection. The CPI figures alone cannot establish the right balance for a municipality or regulator.
For businesses, the useful response is to distinguish fuel-sensitive operating costs from other administered charges when testing margins and negotiating supply terms. A distributor and an office-based firm need not face the same exposure. Consumer CPI is not a business-input index and supplies no automatic justification for a 13% price rise.
Households remain affected, but not uniformly. Fuel pressure can reach people without cars through transport charges. Who ultimately bears an input increase depends on whether suppliers absorb it, pass it on or reduce another expense.
Risks and limitations
These are national aggregate measures, not a municipal tariff audit or a finding about any individual household. A lower inflation rate still generally means prices are rising, only more slowly. Fuel sits within the administered-price measure, so adding the two rates would double-count overlapping categories. None of these figures establishes profiteering, tariff legality or the appropriate price increase for a particular firm.
What happens next?
Stats SA lists 18 November 2026 as the expected release date for October CPI. The next test is whether the gap narrows because fuel pressure eases, because other administered inflation slows, or because prices elsewhere accelerate. Those outcomes would have different implications. Until then, September remains a backward-looking measurement, not an October forecast.
SBL bottom line
September supports a narrower conclusion than “all regulated prices are surging”. Fuel changes the story about acceleration, while the non-fuel administered basket still shows substantial annual pressure. Sound scrutiny should keep both facts in view and examine the funding and cost basis of particular charges separately.
Sources and methodology
Primary evidence: Stats SA P0141 August 2026, released 23 September 2026; P0141 September 2026 and its explanatory article, released 21 October 2026; DMPR statement published 31 August 2026, effective 2 September. Links appear above. The administered-price definition is from the September release glossary. Table changes equal September’s annual rate minus August’s. Gaps: 10.6 − 3.3 = 7.3 and 13.0 − 3.3 = 9.7 percentage points. No missing values have been estimated.
Prepared and last source-checked: 26 October 2026. Draft for editorial review; not yet published.
[…] reading: SBL’s administered-price inflation analysis examines a separate price measure and why its components […]