The geography of UK crime reporting.

A high count on a postcode report is sometimes a high-crime area and sometimes a high-reporting area. The two look identical on the page — a number is a number — but they mean different things, and a reader who treats them as interchangeable will draw the wrong conclusion about half the time. The distinction is not philosophical. It changes what the figure says about a place and what a sensible comparison looks like. This piece is about the gap between recorded crime and experienced crime, and the systematic ways that gap varies by geography. Some of the variation is real; a meaningful share is an artefact of how, where, and by whom incidents enter the recording system.

Recorded crime vs experienced crime

There are two yardsticks for measuring crime in England and Wales, and they tell different stories because they measure different things. The first is police-recorded crime, which is what Police UK publishes and what every postcode tool ultimately draws from. The second is the Crime Survey for England and Wales, a household survey run by the Office for National Statistics that asks a representative sample of adults what they have actually experienced in the previous year, regardless of whether they reported it.

Police-recorded crime depends on a chain of events. Someone has to notice an incident, decide it is worth reporting, and the police have to receive the report, agree it meets the threshold under the National Crime Recording Standard, and log it under the correct category. Each link has its own attrition rate. The Crime Survey skips most of those links by asking the victim directly. The gap between the two — the so-called dark figure of crime — is not a small correction. For some categories, particularly sexual offences, low-value theft, and domestic abuse, the survey figure runs several times higher than the recorded figure. The point is not that one is right and one is wrong. They are answering different questions.

Why urban areas look worse

Three things stack up at the dense end of the spectrum, and all of them push the recorded figure up. The first is density of opportunity: more people in close quarters means more targets, more witnesses, and more incidents anyone is in a position to notice. The second is density of reporting infrastructure. Town and city centres carry more CCTV, more foot patrols, and faster response times — all of which raise the share of incidents that turn into a recorded crime. The third is simply population — a half-mile circle in central Birmingham contains tens of thousands more residents, commuters, and visitors than the same circle in rural Powys, and a per-circle figure that ignores that scale will of course look enormous.

None of this means an urban postcode is "more dangerous" in any sense the resident would recognise. It means more of what happens there is captured by the recording system. A useful test, when staring at a high urban figure, is to ask how many of those incidents the same area would have generated under a rural reporting regime — typically a much smaller number, with the gap doing more of the work than the behaviour itself.

Why some rural areas look mysteriously safe

The mirror image is rural under-reporting, and it is the dominant effect at the sparse end of the map. The nearest station may be twenty miles away. The threshold for picking up the phone for a non-urgent matter is higher when the response will take an hour to arrive. Community policing footprints are smaller. A meaningful share of disputes are resolved informally — a word with the farmer next door, a complaint passed through the parish council — that would, in a town, end up as a 101 call and a recorded ASB incident.

The Crime Survey makes the size of this effect visible. Victimisation rates for rural respondents do not sit at the floor the recorded figures imply; they tend to land closer to the urban figure than the published counts would suggest. The gap is reporting rather than victimisation. A postcode that shows almost nothing on the map is sometimes genuinely quiet and sometimes a place where the same incidents happen but never make it into the data.

The "hot street" effect

Even within a single postcode, the geography is rarely uniform. A surprising share of a residential area's recorded total can come from one or two specific land uses inside the radius. A 24-hour Tesco Extra in an otherwise quiet suburb can account for a disproportionate share of shoplifting. A retail park on the edge of the circle pulls vehicle-crime figures up. A nightclub strip half a mile from a row of family homes can make the public order line look alarming. A transport hub concentrates theft from the person into a single dense pixel.

The implication is straightforward. When the headline number looks high, the category breakdown is usually the next thing to look at, and the geography behind it the thing after that. A postcode total is one number, but the streets that produced it almost never contribute equally. Asking whether a single car park, retail unit, or licensed venue is doing most of the work often pulls a frightening total back down to something specific and bounded.

The category gradient

Different categories distribute differently across the urban-rural spectrum, and the shape is worth carrying in the back of the mind. Burglary tracks property density and access patterns. Vehicle crime tracks parked-car density almost mechanically — long stretches of unbroken kerbside parking generate more of it than driveways and garages. Anti-social behaviour tracks reporting infrastructure as much as actual nuisance, which is why suburban ASB counts can outpace rural counts even where the underlying levels of disturbance are comparable. Violence and sexual offences are weighted toward urban centres, but a meaningful part of that gap is reporting too — urban victims are more likely to report, and urban witnesses are more likely to be present at all.

Knowing the rough gradient helps the reader judge whether a postcode's profile is unusual or simply typical for its type. A city-centre postcode with high public order and low burglary is doing what city-centre postcodes do. A rural postcode with low everything is doing what rural postcodes do on the recorded data — which is not the same as saying nothing is happening there.

Practical reading — three rules

Three habits make most of the difference. The first is to compare like with like. A half-mile circle in central Manchester and a half-mile circle in the Yorkshire Dales are not measuring the same population, the same reporting regime, or the same set of land uses, and concluding that one is "ten times more dangerous" on the strength of a ratio is a category error. A useful comparison is between two postcodes of broadly similar type — two suburbs of the same city, two market towns of similar size.

The second is to read the category breakdown whenever the headline figure looks alarming. A total dominated by ASB and shoplifting is geometrically different from one dominated by burglary and violence, and the breakdown is usually enough to tell which kind of place produced the number. A high count built almost entirely from one category is often traceable to one or two locations inside the radius.

The third is to apply the same scepticism to suspiciously low figures as to suspiciously high ones. An area that shows almost nothing might be quiet, or it might be under-reported. Both possibilities deserve a moment's thought. The recorded figure is the best routinely-available signal we have, but it is a signal, not the thing itself, and the geography it draws is shaped at every turn by who is doing the reporting and where.