There can be a gap between a price displayed on a shopping website and the price paid at a market. For the Bank of Ghana, this difference lies at the heart of a practical question: how can price pressures be detected earlier without confusing online trade with the full range of household spending?
Researchers at the central bank are proposing an approach to complement conventional price surveys. Their tool, called “e-Inflation”, uses prices published online to produce more frequent estimates. At this stage, it does not replace the official index calculated by the Ghana Statistical Service (GSS), the national statistics agency.
A faster measure, not a new official index
The project addresses a familiar challenge for economic institutions: monetary policy decisions rely on data that sometimes describe a situation that has already passed. In Ghana, official inflation figures for a given month are published about two weeks after it ends. Automation could provide interim signals while the authorities await the consolidated series.
This question is of direct relevance to the Bank of Ghana and its Monetary Policy Committee. Inflation, or the rate at which prices rise, is part of the analysis that guides interest rate decisions. These decisions affect the cost of credit and financing conditions for businesses and households. A faster indicator could therefore inform the debate, without determining monetary policy on its own.
In its study published on 23 October 2026, the Bank of Ghana describes a method that automatically collects prices from 120 websites. The researchers then match the data to categories in the official consumer basket. This basket contains 307 goods and services. The GSS’s traditional method, by contrast, is based on nearly 47,900 price quotations collected in 57 markets.
The volume of online data amounts to about 500,000 price observations. This does not mean the tool covers 500,000 different products: the observations are aggregated to estimate changes across the 307 items in the basket. The Bank of Ghana’s presentation of the study explains that the system is intended to produce estimates of headline, food and non-food inflation.
Results are close, but limitations remain to be assessed
To test the method, the authors compared its estimates with official figures between May 2021 and December 2023. They report a strong correlation between the two series: 0.9941 for headline inflation, 0.9836 for food and 0.9944 for other goods and services. These coefficients describe how closely the observed trends track each other. On their own, they do not prove that every online price accurately represents the price paid by all consumers.
The Ghana News Agency, the national news agency, also reported on the study on 26 October. It noted that the tool is presented as a way to obtain indications before official statistics are published, not as a substitute for them. The distinction matters: GSS figures remain the public benchmark for measuring inflation in Ghana.
Comparison with GSS price surveys provides a useful initial test, but does not settle the question of representativeness. A price displayed online may not be up to date, and not all goods are sold on the internet. Purchases made at physical markets, neighbourhood shops or through informal channels may be less visible in data collected from websites. Households that buy mainly offline could therefore be less well represented.
The method also depends on the quality of the selected websites: the availability of prices, the regularity of updates and the ability to collect the data. Changes in the number of accessible platforms or in their practices could alter the sample. Results will therefore need to be checked regularly against prices collected on the ground, including in different regions and sales channels.
From statistical signal to effects on households
For monetary policymakers, the potential benefit is the time saved. A more frequent series could help identify a turning point or an acceleration between monthly releases. But a rapid signal is useful only if its biases are understood. An estimate that tracks the national series closely does not automatically reveal the difficulties faced by a particular group, region or type of retailer.
For households, inflation translates into very concrete trade-offs between food, transport, housing and other expenses. Lower-income households often spend a large share of their budgets on essentials; their experience of prices also depends on where they shop. An indicator built from online shops should therefore be read as a macroeconomic complement, not as a direct measure of the cost of living for every household.
Businesses, too, can track changes in costs and anticipate decisions on procurement or pricing. However, decisions based on an experimental indicator would carry risks if it diverged from prices actually observed in local sales channels. Transparency about the products covered, the websites selected and the adjustments made will therefore be crucial.
A regional issue too
Ghana is a member of the Economic Community of West African States (ECOWAS), which brings together countries facing shared challenges in monitoring prices and the movement of goods. A validated Ghanaian experiment could contribute to regional discussions on digital statistics. It would not, however, be automatically transferable: access to online trade, shopping habits, consumer baskets and markets vary from one country to another.
The next step will therefore be less technological than institutional: establishing whether the tool remains reliable over time and how it can be compared with GSS surveys. The researchers present their work as a study, not as an announcement of an official change in methodology. For the Bank of Ghana, the challenge is to save time without losing sight of the markets and consumers that digital data capture poorly. Only then can a faster indicator strengthen, rather than cloud, the understanding of Ghana’s economy.



