Nine Hunter councils logged 38 code of conduct complaints and found no breach. The published data cannot tell you what that means

The NSW Office of Local Government publishes a workbook of council performance data every year. We have been through it twice before, for the seven financial ratios it stopped collecting and for what each Hunter council owes on its infrastructure backlog. Three of its 124 columns cover the code of conduct: complaints received, the cost of dealing with them, and the number of finalised complaints that were investigated and found to be a breach. Nobody appears to have put the Hunter ones side by side.

What the nine councils reported

Source: NSW Office of Local Government, Your Council time series data 2024-25, Councils sheet, columns 111 to 113. The rows are as published. Upper Hunter reports two complaints and no cost, which is one of eight councils statewide where those two columns do not reconcile.

The figure that stands out is not a Hunter one at first glance. Newcastle spent $62,689, which is the sixth highest of any council in NSW, on the seventeenth highest number of complaints. Only five councils in the state spent more, and three of them logged more than twice as many complaints. On our arithmetic that is $6,965 per complaint against a state average of $2,436, and none of the nine resulted in a recorded breach.

That division is ours, not the Office of Local Government's, and it is a blunt instrument: the reported cost includes staff time, so a council that runs complaints through its own governance team and one that refers everything to an external conduct reviewer will produce very different numbers for the same work. What it does establish is that the money is not tracking the volume.

Statewide, 692 complaints produced 99 findings, and 74 of them are one council

Across the 127 councils that reported, the totals are 692 complaints, $1,685,874, and 99 finalised investigated complaints where a breach was found. That last number looks healthier than it is. Snowy Monaro Regional Council accounts for 74 of the 99, and reports 74 breaches from 74 complaints, every single one. A council where every complaint received is also a breach found is almost certainly counting something different from its neighbours, and we would not read it as a finding about Snowy Monaro. Set it aside and the rest of NSW found 25 breaches across 618 complaints, about 4 per cent.

At the other end, Liverpool City Council recorded 148 complaints and spent $311,897, both the highest in the state, and recorded no breach. A year earlier it recorded five complaints and $15,000. We cannot tell you what happened there from this dataset, and that is the point of this article rather than an aside.

The statewide totals moved sharply year on year. In 2023-24 the same columns carried 381 complaints and $1,415,214. So complaints are up about 82 per cent and spending up about 19 per cent, and rather more than half the increase in complaints is Liverpool and Snowy Monaro between them.

Why a complaint so rarely becomes a breach, in the procedures' own words

The low conversion is not evidence of a whitewash. It is close to what the process is designed to produce, and the Procedures for the Administration of the Model Code of Conduct say so directly.

A conduct reviewer may only decide to investigate a complaint where satisfied that it is a code of conduct complaint, that the matter could not or should not be resolved by other means, and that the alleged conduct is sufficiently serious to warrant the formal censure of a councillor under section 440G of the LGA or disciplinary action against the general manager under their contract of employment if it were to be proven. That is a high bar, set before anything is investigated at all.

And where a complaint is resolved without investigation, the procedures are explicit that this is not a verdict: The resolution of a code of conduct complaint under clause 6.13, paragraphs (b) or (c) is not to be taken as a determination that there has been a breach of the council's code of conduct.

So a council can receive a dozen complaints, deal with every one properly, and correctly report zero breaches. Read the other way, a council can receive a dozen complaints, decline them all at the first gate, and report exactly the same zero. The published data gives a reader no way to tell those two councils apart.

The two numbers that would tell them apart are collected and not published

This is the finding. Clause 11.1 of the procedures sets out the statistics a council's complaints coordinator must report, and the list is longer than what reaches the public workbook. It includes the number of code of conduct complaints finalised by a conduct reviewer at the preliminary assessment stage during the reporting period and the outcome of those complaints and the number of code of conduct complaints investigated by a conduct reviewer during the reporting period, along with the outcome of completed investigations and the number of matters reviewed by the Office.

Clause 11.2 requires the council to give the Office a report containing those statistics. The published workbook carries the total received, the total cost, and a breach count. The two statistics that convert a complaint count into something a resident can interpret, how many were finalised at the first gate and how many were actually investigated, are collected from every council in the state and appear nowhere in the public file.

Put plainly: the Office of Local Government holds the denominator and publishes only the numerator.

Three other things the file does not line up

The reporting period does not match the file. Clause 11.1 sets the reporting period as the year to September. The workbook is a 2024-25 file and every financial column in it is a financial year. A reader comparing a council's complaint count against its spending in the same row is very likely comparing two different twelve-month windows, and nothing in the workbook says so.

The statutory statistics cover councillors and the general manager only. Clause 11.1 is headed reporting on complaints about councillors and the general manager, and every limb of it is framed that way. The workbook column is labelled simply Number of Code of Conduct Complaints Received, with no scope stated. If the published figure is the clause 11.1 figure, it is not a count of complaints about a council, it is a count of complaints about its elected members and its most senior employee. We have not been able to confirm the scope from any published note, and have not sought comment from the Office.

One number in the file is impossible on its face. In the same workbook, Cessnock City Council's Councillor Expenses figure is published as minus $398,000. It is the only negative value in that column anywhere in the state, and there were none at all in the 2023-24 file. A negative expense is not a finding about Cessnock and we do not report it as one; it is a data quality flag on a file that residents are invited to compare councils with. Alongside it, seven councils report complaints with a cost of zero, and one reports a cost with no complaints.

What this is worth to a resident, and what it is not

The Office of Local Government publishes its own caution beside this data, and it is a fair one: key performance indicators, when compared in isolation, do not give the full picture of a council's performance, and they don't explain why discrepancies occur. That is true of these three columns more than most, because two of the three are inputs and the third is an outcome measured after a filter the file does not describe.

What the numbers do support is narrow and still useful. Nine Hunter councils logged 38 code of conduct complaints about their councillors and general managers in a year and spent $106,906 on them. Three logged none. Newcastle spent more than all but five councils in NSW. No breach was recorded anywhere in the region.

What they do not support is any claim that Hunter councils behave better than others, or that their complaint processes are soft. Both readings fit the same published figures equally well, which is the strongest argument for publishing the two columns that would separate them. Every council already sends them in.