Inflation Slowed. Your Bills Did Not: Malta’s Cost-of-Living Test

The number fell. The price level did not.

You notice the change at the till long before a statistical release explains it. When Malta’s inflation rate falls, prices are still generally rising; they are simply rising more slowly than before. The NSO defines the annual inflation rate as the change between a reference month and the same month a year earlier. That is a rate of change, not a promise that prices have returned to their old level.

In August 2026, Malta’s HICP inflation rate was 2.0%, down from 2.1% in July. Its 12-month moving average was 2.3%. The RPI rate was 2.6%, down from 2.7%, with a 12-month moving average of 2.6%. Those are signs of disinflation: a slower increase in the price level. They are not evidence that the price level has fallen.

How much higher are prices than before the surge?

A more useful question for a household budget is often: how far has the price level moved since the shock years? Using December-to-December official rates, the HICP rose by approximately 16.0% between December 2021 and December 2025. The calculation compounds the annual changes of 7.3% in 2022, 3.7% in 2023, 1.8% in 2024 and 2.4% in 2025: (1.073 × 1.037 × 1.018 × 1.024) − 1. This is a transparent all-items calculation, not an annual-average comparison.

The comparable RPI calculation gives approximately 15.6% over the same December 2021-to-December 2025 period, using annual rates of 7.35%, 3.6%, 1.2% and 2.7%. The small difference is expected: RPI and HICP are built differently. Neither figure says that every household’s shopping basket rose by exactly that amount, but both show why a current rate near 2% should not be read as a return to 2021 prices.

There is also a methodological boundary worth keeping visible. The cumulative HICP illustration ends in December 2025 and uses the series then in force. From the January 2026 release, NSO moved HICP to ECOICOP version 2 and rebased the reference period to 2025=100. That change does not invalidate the earlier December-to-December calculation, but it is another reason to avoid presenting index comparisons as more exact than they are.

Food, housing and services still shape the pressure

Food remains one of the clearest reasons a lower headline rate may not feel reassuring. In December 2025, food and non-alcoholic beverages recorded HICP inflation of 3.7% and made the largest upward contribution to the overall HICP rate, at 0.68 percentage points. In August 2026, food was also the largest contributor to RPI inflation, at 0.55 percentage points. The two food categories are not identical: the RPI Food Index includes restaurant services and take-aways.

Housing is another important distinction between a national headline and a household bill. In August 2026, the RPI Housing Index rose 6.8% year on year, faster than the all-items RPI. Within it, rent rose 3.5%, materials for house maintenance 8.0% and services for house maintenance 6.7%. These are index movements: they do not mean that every tenant, owner or maintenance customer faced those exact increases.

Energy-related prices were comparatively stable in the August RPI release: water, electricity, gas and fuels recorded 0.0% annual inflation. That can matter greatly to households, but it does not freeze the rest of the budget. In the August HICP release, restaurants and accommodation services contributed 0.41 percentage points to overall inflation, while housing, water, electricity, gas and other fuels and transport each contributed 0.28 points.

Why RPI and HICP do not give the same answer

RPI is the more direct domestic household lens, while HICP is the harmonised measure used for comparisons across the European Union. The NSO says RPI covers private households only. HICP also covers institutional households and foreign visitors. RPI uses a fixed base with periodically reviewed weights; HICP is chain-linked and can introduce newly significant goods and services each year.

The published 2025 weights illustrate the difference. HICP assigned 181.67 per thousand to food and non-alcoholic beverages and 102.16 per thousand to housing, water, electricity, gas and other fuels. RPI assigned 21.49% to Food, 7.90% to Housing and 3.31% to water, electricity, gas and fuels. These are national expenditure weights, not a recommended family budget or a hardship threshold. A household that spends unusually heavily on rent or food can therefore experience a different inflation rate from the national average.

Who can the average miss?

An index is designed to describe broad price movement, not to reproduce an individual’s accounts. A renter may be more exposed to housing costs than the national weight suggests. A pensioner may spend a larger share on food, utilities or health-related services. A lower-paid worker has less room to absorb a fixed increase in essentials, while a family with children may face a different mix of food, transport, education and housing costs. The reviewed NSO material supports these limits, but it does not provide a household-specific inflation rate for each group.

This is why a headline such as “inflation is only 2%” can be technically correct and still incomplete. The rate describes the latest annual movement in a weighted basket. It does not tell you whether your rent was renewed, whether your usual groceries rose more quickly, or whether your income has enough flexibility left after essentials. National averages are useful for tracking the economy; they are not household ledgers.

Have wages kept pace? The official data need care

NSO Labour Force Survey data show average monthly basic salary rising from €2,063 in Q1 2025 to €2,270 in Q1 2026, an increase of approximately 10.0%. But this is a nominal, pre-tax measure. It excludes overtime, allowances and bonuses, and the figures are provisional and subject to revision. It is therefore not safe to treat the 10.0% change as a confirmed gain in every worker’s purchasing power.

A proper real-earnings comparison needs compatible earnings and price coverage, a clearly matched period and an appropriate deflator. The available LFS salary measure is not a matched household-income measure: it excludes self-employed income, tax, social contributions and non-basic pay. Nor does an average salary describe the worker at either end of the distribution. In Q1 2026, the reported average monthly basic salary ranged from €1,410 for elementary occupations to €3,628 for managers.

What Budget 2027 should be judged against

For households, the relevant baseline is not only the latest inflation rate. It is the accumulated price level since the surge, the purchasing power of income after tax and contributions, and the particular costs that dominate each household’s budget. For the government, the same discipline means separating an observed price statistic from a future policy announcement. As at the 4 October 2026 evidence cutoff, the pre-budget evidence does not establish the final Budget 2027 package or the final 2027 COLA amount.

One important qualification belongs here. The 2023 national minimum-wage agreement and the implementing wage instruments already set a scheduled basic national minimum wage of €210.73 per week from 1 January 2027 for employees aged 18 and over, before the statutory COLA awards for 2024, 2025, 2026 and 2027. The figure is therefore part of an established roadmap, not a new Budget 2027 announcement. The final 2027 COLA amount was not established at the evidence cutoff, and sectoral wage regulation orders may apply in particular activities.

The practical test is straightforward. Has the measure lowered the price level, protected income against the price level, or merely slowed the next increase? Does it reach renters as well as owners, pensioners as well as employees, and households whose essentials take up most of their income? Those questions belong alongside the headline rate. A slower climb is welcome; it is not the same thing as reversing the climb.

Two official measures, two useful lenses