The only two offences rising in Wollongong are both domestic violence. Across NSW the same measure is flat

Read the caution before the numbers, because it is not boilerplate. These are recorded incidents, meaning things police wrote down. BOCSAR is blunt about what that does and does not tell you: recorded crime statistics for some categories DO NOT accurately reflect the actual level of crime in the community, because the count is affected by public willingness to report crime and by shifts in policing policy. It adds that irrespective of how trends in recorded crime are presented, their interpretation is a difficult task. For domestic violence in particular, a rise in recorded incidents is consistent with more violence, with more of it being reported, with more of it being recorded once reported, or with any mixture of the three. We cannot tell you which, and neither can this dataset.

What the table says

BOCSAR reports each offence as either a significant percentage trend or as Stable, which means its test found no significant trend rather than that nothing changed. Here is Wollongong against the state, in incidents and in rate per 100,000 people, for the twelve months to March in each year.

Two things the rates change about the story

First, Wollongong is not an outlier on level. Its domestic violence assault rate of 407.0 per 100,000 sits below the NSW rate of 451.9, and BOCSAR ranks it 73rd of the LGAs it ranks, with a ratio to the state rate of 0.9. The finding here is about movement, not about Wollongong being unusually violent.

Second, the breach figure is the one where Wollongong sits above the state, at 393.4 per 100,000 against 331.0, a ratio of 1.2. And it is rising two and a half times faster than the state figure. An apprehended violence order is a court order made to protect someone; a breach is recorded when that order is not obeyed. That measure moving is a different signal from an assault count moving, because it counts what happens after the system has already been asked to intervene.

It is a Wollongong finding, not an Illawarra one

We downloaded the tables for all three councils this masthead covers, and the honest result is that the pattern does not extend past the city.

• Shellharbour: domestic violence related assault is Stable on BOCSAR's test, at 303 incidents and a rate of 371.5, below the state rate. Breach AVO is also stable. The only offence flagged as moving is other theft.
• Kiama: domestic violence related assault is Stable, and the count actually fell, from 31 to 28. The rate, 120.8, is roughly a quarter of the state rate. Breach AVO numbers are too small for BOCSAR to calculate a trend at all.

Small numbers are why Kiama carries no trend, and that is a feature of the method rather than a gap in it: BOCSAR will not compute a trend it cannot stand behind, and neither will we. If we had reported this as an Illawarra-wide rise we would have been wrong on two of the three councils.

What we are not saying

We are not saying the Illawarra has become more dangerous, that police are failing, or that any particular policy caused this. None of those claims is available from a table of recorded incidents, and we have not sought comment for a piece that is an arithmetic reading of a public dataset. What we can say is narrow and, we think, worth publishing: on the state's own numbers, using the state's own trend test, the only two offences moving upward in Wollongong are both domestic violence, and both are moving faster than the state.

The obvious next question, whether more orders are being made as well as breached, is not answerable from this table. It is the thing we will go looking for next.