Investigation

A water subsidy line grew for three years without a matching reconciliation

District meter records and the authority's own quarterly filings diverge on how many households the subsidy actually reached.

By Danial Rashid · 13 July 2026 · 6 min read

Case names, entities and officials in this report are composite and illustrative. References to public bodies describe their statutory role only.

A subsidy line that grew

The Sungai Petani District Water Authority's low-income tariff subsidy budget rose from 4.1 million ringgit in the 2023 financial year to 7.8 million ringgit in 2025, a near-doubling over two cycles that the authority's own budget notes attribute to population growth and expanded eligibility criteria. Household registration figures published alongside each year's budget rose more slowly, from 18,400 households in 2023 to 21,900 in 2025 — an increase of nineteen percent against a subsidy line that grew by ninety percent.

Budget lines and registration counts are prepared by different units within the authority on different timetables, which the authority's finance office has cited in past annual reports as a routine source of timing gaps between the two figures. This report sets that explanation against the reconciliation the authority itself produced for 2025, and against the metering data that underlies both.

The authority's finance office has offered the timing explanation in three consecutive annual reports without publishing the reconciling calculation that would let a reader test it. A budget line and a registration count measuring the same underlying population should, over a long enough window, converge rather than diverge further each year, and the widening gap between the two figures is the starting point for the rest of this report.

Reading the reconciliation

A subsidy reconciliation is the document that is supposed to close the gap: it takes the budget allocation, matches it against verified meter readings for registered low-income households, and reports the variance. The authority's 2025 reconciliation, released in response to a routine disclosure request, reports a variance of 2.3 percent between allocated and verified subsidy spend — well within the five percent tolerance the authority's finance manual sets as acceptable.

That reconciliation, however, verifies spend against estimated consumption for roughly forty percent of registered households, not against actual meter readings. The authority's own footnote to the reconciliation states that meter access could not be confirmed for the remaining households within the reporting window, and that estimated figures, calculated from the prior year's readings plus an inflation factor, were substituted. The 2.3 percent variance therefore compares a real allocation against a partly estimated consumption figure, not against confirmed usage.

The five percent tolerance itself is not unusual; utility regulators across the region generally allow a similar margin to absorb rounding, late meter reads and minor billing corrections. What makes the 2025 figure worth examining is not the tolerance but what sits behind it — a reconciliation that reports comfortably within tolerance while relying on estimation for a large share of the households it is meant to verify.

Meters, estimates and the gap

Kilat Meter Services Sdn Bhd holds the district's metering and reading contract, renewed in 2024 for a further three years. Its service log, filed quarterly with the authority, records a completed-reading rate of 91 percent for the district as a whole in 2025 — a figure that sits uneasily against the reconciliation's finding that forty percent of subsidy households specifically could not be confirmed. The gap suggests that unread meters cluster disproportionately among subsidy-eligible households rather than being spread evenly across the district.

Kilat's service log does not break down the completed-reading rate by tariff category, so the clustering described above cannot be confirmed from the documents released so far; it is inferred from comparing two figures the authority and its contractor each publish for different purposes. A reader with access to the raw meter-route data, which the authority has stated it does not publish, could confirm or rule out the clustering directly.

Kilat's contract, reviewed for this report, sets a minimum completed-reading rate of 90 percent as a service standard, with financial penalties for shortfalls measured at the district level rather than by tariff category. A contractor can meet its district-wide target while still under-reading a specific subset of households, such as subsidy recipients, without breaching any measurable term of its agreement — a structural feature of the contract rather than evidence of any particular reading pattern.

What the authority reported upward

The authority's quarterly submissions to the state water regulator report subsidy coverage as a single aggregate figure — the percentage of registered low-income households receiving the subsidy in a given quarter — without distinguishing between confirmed and estimated coverage. Every quarterly submission reviewed for this report states coverage at or above 95 percent, a figure drawn from the same reconciliation process described above rather than from a separate confirmed count.

The regulator's reporting template, published on its own site, asks authorities to report coverage as a single number and does not currently require a breakdown between confirmed and estimated households. Sungai Petani's practice is consistent with the template it was given; the template itself does not distinguish between the two categories of data this report treats as materially different.

A coverage figure reported upward without a confirmed-estimated breakdown is not unique to Sungai Petani; several district authorities operating under the same regulator template report coverage the same way, according to a sample of five other districts' quarterly filings reviewed for context. The pattern appears to be a feature of how the reporting template is built rather than a practice specific to this authority, though Sungai Petani's subsidy line has grown faster than any of the five comparison districts over the same period.

Two versions of the same quarter

For the third quarter of 2025, this report compared the authority's regulator submission, its internal reconciliation working file, and Kilat's service log side by side. The regulator submission states coverage of 96.4 percent. The internal working file, which breaks the figure down, shows 58.1 percent of households confirmed by direct meter reading and the remainder carried forward at the estimated rate described above. Kilat's log for the same quarter shows a completed-reading rate of 89 percent district-wide — below the 90 percent minimum the contract sets as a service standard, though that standard is assessed against the full year rather than any single quarter, and the district's 2025 annual figure of 91 percent meets it.

None of the three figures is inconsistent with the authority's own published methodology; each measures a slightly different thing, and the authority has not published a document that reconciles all three at once. The absence of that single reconciling document, rather than any one figure in isolation, is the gap this report identifies. A subsidy budget that has grown by ninety percent over two years, and a coverage claim built partly on estimation, are each independently unremarkable; together, without a document that ties consumption, registration and confirmed metering into one figure, they are difficult for an outside reader to check.

The authority has confirmed receipt of a further disclosure request seeking the raw meter-route data referenced above and the underlying working papers for the 2025 reconciliation. Both remain outstanding at the time of publication, and this report will note any material the authority subsequently releases.

A coverage figure that blends confirmed readings with a rolled-forward estimate is not wrong, but it is not the same claim as a confirmed figure, and readers deserve to know which one they are looking at.