Skip to main content
Methodology and Notes

How we build the Middle-Income Inflation Index

The full canonical reference for the Deciles 4–7 inflation index, sub-indices, weight structure, Stats SA survey schedule, assumptions and limitations.

Section 1

Data Sources

Price Data. All underlying price indices — including headline CPI and its components (food, transport, fuel, housing, etc.) — are sourced from Statistics South Africa's monthly Consumer Price Index releases (P0141).

Fuel Price Data. Pump price data (diesel 500ppm, petrol 93 and 95 unleaded) sourced from the Automobile Association's monthly fuel price tracker, alongside National Treasury communiqués on fuel levy relief measures.

Expenditure Weights. Weights for constructing the Deciles 4–7 index are derived from Stats SA's official "CPI per expenditure decile" data, based on the 2022/23 Income and Expenditure Survey (IES). For this analysis, weights for deciles 4, 5, 6, and 7 were aggregated to represent the middle-income group.

Section 2

Index Construction

2.1  The middle-income (Deciles 4–7) CPI

The middle-income (Deciles 4–7) CPI is calculated by reweighting the official CPI component indices according to the expenditure patterns of the middle-income group. The year-on-year inflation rate is derived as follows:

$$ \pi_{t}^{\,D4\text{-}7} \;=\; \sum_{c=1}^{n} \left[\, w_{c} \times \left( \frac{\mathrm{CPI}_{c,t}}{\mathrm{CPI}_{c,t-12}} - 1 \right) \times 100 \,\right] $$
$w_{c}$
expenditure weight of category c for Deciles 4–7
$\mathrm{CPI}_{c,t}$
price index for category c in month t
$\mathrm{CPI}_{c,t-12}$
price index for the same category 12 months earlier

These weights reflect each category's share of aggregate household expenditure for the middle-income group, ensuring the index is representative of their consumption patterns.

2.2  Sub-Indices

To analyse distributional price pressures, three subgroup indices were constructed:

CNI
Core Necessities
66.1%

Food & Non-Alcoholic Beverages, Housing & Utilities, Transport.

SEI
Semi-Essentials
7.0%

Health, Education, Insurance.

NEI
Non‑Essentials
26.9%

All remaining categories.

Each sub-index is calculated as a weighted average of its constituent categories' inflation rates, using their within-group expenditure shares.

Section 3

Stats SA's Expenditure Deciles

Stats SA divides South Africa's urban population into 10 expenditure deciles, each representing 10% of households ranked by total monthly expenditure. This classification allows for detailed analysis of how price changes affect households across the income spectrum.

The table below shows the official CPI weights (as a percentage of the national CPI basket) for each decile, based on the 2023 weight update. The four middle deciles (D4–D7) — our middle-income focus — are highlighted.

Decile Monthly expenditure band CPI weight
D1 Up to R35 864 0.71%
D2 R35 865 – R51 168 1.31%
D3 R51 169 – R64 855 1.85%
D4 Middle R64 856 – R80 320 2.54%
D5 Middle R80 321 – R97 976 3.30%
D6 Middle R97 977 – R120 028 4.42%
D7 Middle R120 029 – R150 774 6.23%
D8 R150 775 – R202 419 8.89%
D9 R202 420 – R307 885 15.03%
D10 R307 886 and more 55.72%

Source: Statistics South Africa, CPI 2023 Weights (P0141.5).

The Deciles 4–7 group — comprising the four middle expenditure deciles — represents the core of South Africa's urban middle class. These households fall between the bottom 30% and top 30% of the expenditure distribution, capturing what is often described as the "lower-middle" to "middle-middle" income segment.

By constructing a weighted CPI index for Deciles 4–7, we isolate inflation pressures experienced by this economically significant and policy-relevant group, separate from the often-divergent experiences of the lowest-income and highest-income households.

Section 4

Decile 4–7 Weight Structure

Category shares within the middle-income basket, derived from the 2023 CPI weight update.

Category Weight
Food & Non-Alcoholic Beverages 30.5%
Housing & Utilities 27.3%
Transport 8.3%
Insurance 5.3%
Clothing & Footwear 5.7%
All other categories 22.9%
Total 100.0%
Section 5

Survey Schedule

Stats SA's survey cadence for items that are not collected monthly.

Frequency Survey months Items
Quarterly Jan, Apr, Jul, Oct Gymnasium fees (Recreation & culture); Funeral expenses, funeral policies, gravestone maintenance (Misc)
Feb, May, Aug, Nov Private-sector hospitals (Health); Rugby tickets, Cricket tickets (Recreation)
Mar, Jun, Sep, Dec Actual rentals, owners' equivalent rent (Housing & utilities); Domestic workers' wages (Household contents); Taxi & train fares (Transport); Motor vehicle insurance (Misc)
Biannual Jan, Jul Building insurance, household content insurance (Misc)
Feb, Apr Medical aid (Misc)
Apr, Oct Television licences (Recreation & culture)
May, Nov Electricians, plumbers (Housing & utilities)
Jul, Aug Municipal charges for water, property rates & electricity (Housing & utilities)
Annual January Post box rental (Communication); Lotto tickets (Recreation)
February Private-sector doctors & dentists (Health)
March School & university tuition fees (Education); Toll fees (Transport); University boarding fees (Restaurants & hotels); Crèche fees (Misc)
April Public-sector hospitals & doctors (Health); Stamps (Communication)
Other Jan, Mar, Apr, Jun, Jul, Sep, Oct, Dec Local bus fares (Transport)

Source: Statistics South Africa.

Section 6

Key Assumptions & Limitations

Experimental Index
This constructed index is for analytical purposes only and is not an official statistic published by Stats SA.
Weight Currency
Decile expenditure weights are based on survey data (IES 2022/23). If the survey is not the most recent, weights may not fully reflect current spending patterns.
Within-Group Homogeneity
The index assumes uniform spending patterns within each decile and across the combined middle‑income group, which may not capture the full diversity of individual households.
Price Coverage
The index relies on the same price collection and item coverage as the official CPI, which may not perfectly match the consumption basket of every middle‑income household.
Driver Analysis
Observations about inflation drivers (e.g., housing, food, transport) are based on the relatively higher expenditure weights these categories hold in the decile data.
Second-Round Effects
References to second-round pass-through in Section 4 are based on observed correlations between transport-exposed categories and the timing of the fuel shock. These are analytical observations, not formal econometric pass-through estimates.
Headline CPI Dominance
National headline inflation is dominated by higher‑expenditure deciles, which carry larger weights in the aggregate basket. This can lead to divergence between middle‑income inflation and the official headline rate.
Section 7

Disclaimer

This analysis and the derived indices are intended for informational and research purposes. They should not be used as a basis for formal policy decisions, wage negotiations, or financial indexation without further validation and expert consultation. The author bears no responsibility for decisions made using this data.

Section 8

References

  • Oosthuizen, M. (2007). Consumer Price Inflation across the Income Distribution in South Africa.
  • Statistics South Africa. Consumer Price Index: July 2026 (Statistical Release P0141). Retrieved from www.statssa.gov.za.
  • Statistics South Africa. Consumer Price Index: 2023 Weights (Statistical Release P0141.5). Retrieved from www.statssa.gov.za.
  • Statistics South Africa. Income and Expenditure Survey (IES) 2022/23, Retrieved from www.statssa.gov.za.
Fresh data available
Updated to