Why nobody seems to know if crime is up or down
Back in 1993, a rapidly rising Tony Blair caught the public's attention by pledging to be "tough on crime, tough on the causes of crime".
Liverpool toddler Jamie Bulger had been murdered and violence would soon peak at around four million incidents a year. Records show the public were really worried about crime.
Another big story that year was the creation of the European Single Market - and nowhere near the same number of people were bothered about that.
How times change. Today, more than half of those polled say the EU is one of the most important issue facing Britain today. Only 8% are concerned by crime.
But the latest release of data from the Office for National Statistics includes some potentially worrying signs ahead of an election where crime is almost certainly not going to figure highly in the minds of voters, let alone the parties after their vote.
For years, the Crime Survey for England and Wales (CSEW), the main and most comprehensive measure of what crime people say they experience, has been charting a long-term fall in offences which is consistent with what's been happening in every other comparable nation.
The CSEW says that traditional crime is broadly static at just over six million offences in the year to December 2016. When you add to that experimental data on cyber and fraud, it comes to 11.5m crimes.
There are holes in the CSEW that are difficult to fill - it doesn't cover all types of victims and all types of offences and only recently began asking people about fraud and computer-related crime. But, broadly speaking, it's considered to be a good measure of the long-term trends.
At the same time, police recorded crime has gone up to 4.8m offences - a rise of 9% - and it has seen rises in both serious violence and "traditional" offences such as burglary and robbery.
Those figures - the data collected from what coppers actually know - have been having a torrid time.
There have been a series of revamps and changes in methodology amid concerns that not all forces are recording accurately what's going on - particularly in relation to violence.
The resulting data still isn't considered to be 100% trust-worthy as a "national statistic".
And just to further confuse the picture, there's an entirely independent and highly regarded measure of violence from hospitals that says it has been falling substantially.
So that's as clear as mud. What's going on?
Some of the 9% increase in police recorded crime is almost certainly down to changes in how things are being recorded, but the ONS also stresses that those tweaks can't account for all of the change
On violence, the CSEW estimates there were 29 incidents per 1,000 people - a 4% rise on last year which may not be statistically significant (there are always blips up and down in data).
Police on the other hand recorded a 19% rise in all violent offences - up to 1.1m incidents over the same year.
While the overwhelming number of violent incidents result in either no or minor physical injury, there are some concerning rises with guns and knives.
Firearms offences went up 13%, although all incidents involving guns are still down almost 50% on a decade ago.
Knife crime has gone up 14% during the last year, returning to levels last seen six years ago when it became a political and policing priority.
Both of these rises have already been concerning chief constables, with the new Metropolitan Police Commissioner Cressida Dick saying tackling these offences may define her tenure.
As for fraud and cyber crimes, there are some more concerning figures here. The crime survey hasn't yet collected two complete years of data on these offences - but statisticians already think that it's a rising trend based on other sources.
There were about 5.4m fraud offences in the year to December, with almost 2m of those being computer related, such as online scams or malware used to trick people into providing access to their online accounts.
The National Fraud Intelligence Bureau recorded a 4% increase in offences, but it's not quite clear what's going on nationwide because not all victims come forward. And, just as seriously, not all police forces are geared up to understand what's going on in their patch.
This raises real strategic resources questions for the police: while your average detective knows a thing or two about gathering fingerprints and tracking burglars - spreadsheets and malware are a totally different world.
Marian Fitzgerald, a leading criminologist at the University of Kent, says crime stats have become "a dog's breakfast".
She argues that not only has the crime survey failed to properly inform the public about the emergence of new types of crimes, the police have been caught massaging figures under pressure from successive governments.
"I think the police figures stabilised after the outcry in 2014 when the police were shown to have been fiddling their figures, particularly in relation to violence and sexual offences," says Professor Fitzgerald.
"Any rises that we are seeing now are genuine. Improvements [to police recording practices] stabilised a long time ago."
So is society becoming more criminal, more dangerous? The ONS is working on ways to assess the impact of crime, including a measure of the severity that different crimes cause: an attempted break-in to a shed to nick a lawn mower is wholly different to an act of violent domestic abuse.
Tom Gash, an author and former Whitehall adviser on crime, says: "If you take a centuries-long perspective, [the figures ] are great news. There have been radical changes in the reduction of violence in modern society.
"You could discount [the most recent figures] and say it is a very short-term thing, but the thing that worries me is that this is a rise in violent crime that emulates the US.
"The crime drop in America preceded the crime drop we saw here.
"The truth is we don't really know what's going on. We're not sure if this has been caused by cuts in policing or mental health or youth services, or some broader societal factor that we haven't yet worked out."