Several months ago I posted on the New York Times' amazement at "wonderfully puzzling" declines in US crime rates.  When "everyone knows" that increases in poverty force people into crime, why is US crime is falling steeply, they asked.  Why is robbery, for example, falling even while unemployment increases and unemployment insurance runs out?  

In fact that link between unemployment and crime that "everyone knows" is simply not established by the facts. I find it hard to accept myself.  I'm sure most of us have ourselves got up to no good more when idle, than when busy. We know there is more crime in poor neighborhoods than in affluent suburbs.  So we assume that if a community becomes more prosperous, crime rates will go down, and if incomes go down, crime rates go up.

The assumption is treated as axiomatic in an interesting post on the Standard. It throws some light on the common claim that crime rates are falling. But the assumption in the first paragraph is probably false.

Police Minister Judith Collins’ announcement that crime dropped last year left more than a few people scratching their heads. The economic conditions, especially high unemployment, should mean more crime, not less. Now, we’re starting to learn the answer: procedural changes that have wiped thousands of crimes off the stats

Three years ago Prof James Q Wilson summarised for the LA Times, what was known then about the connections between poverty and crime. James Q. Wilson is  is the author of "Thinking About Crime," the  coauthor of "Crime and Human Nature" and the co-editor of "Crime: Public Policies for Crime Control." He was the key intellectual whose work underpinned the revolution in US criminal justice policy that largely preceded their huge drop in crime rates.

His book  "The Moral Sense" was influential on me when I was ACT's Justice spokesman.  In 2004 I went to talk to him in Santa Barbara. He answered my questions about three strikes and other criminal justice matters but his interests had moved on. He told me he was pre-occupied with understanding the Moslem world, working out how a liberal state might cope with a growing illiberal religion.

The LA Times piece shows he had not abandoned criminal justice policy entirely. It summarised economic studies showing that the popular assumption (connecting crime rates and poverty) is sometimes true, but not always.

For example, up to 2008 it was safe to conclude that  the US burglary rate goes up by 2 percentage points for every 1-percentage-point increase in the unemployment rate. So if the unemployment rate rises from 6 percent to 8 percent (which is about what it was in California then), the burglary rate would increase by 4 percent.  Because burglaries aren't measured accurately (some are never reported, and police vary in how they report the statistics) it's not certain that people would even notice so small an increase.

 A lot of other factors affect the crime rate as well. It often goes up when the population gets younger, and when drug abuse becomes more common. Murder rates are profoundly influenced, at least in big cities, by gang activity.

Wilson emphasized that studies don't necessarily tell us why gang activity changes, though changes in behavior can be heavily influenced by what the police do, whether gang truces have worked and whether gangs are fighting over drug and other illegal transactions.

 Such uncertainties make it hard to be sure why crime rates rise and fall:

•  In the 1960s, the US national homicide rate rose by 43 percent, even though the country was in a period of great prosperity and low unemployment.
•  The homicide rate fell in the 1980s, even as the economy was wobbling, with high interest rates and a steep rise in business bankruptcies.
•  In the 1990s, the murder rate fell by 39 percent at a time when unemployment also was declining.

So can the economy help explain fluctuations in crime?  Sometimes yes, sometimes no, says Wilson.  It would be difficult to link rising crime during the prosperous 1960s to economics.  On the other hand, a declining economy provides plausible theory to explain increases in crime during the 1990s.  Yet some rates began to drop significantly, even ahead of President Clinton's reforms in 1996. Matters become even more complicated if one goes back to the Depression of the 1930s.  There was no FBI data on crime rates at that time, but several studies of individual cities suggest that crime rates fell even though one-quarter  of all Americans were unemployed.  Why?  One reasonable hypothesis is that the Depression pulled families together, and this cohesion inhibited crime.

 Wilson does not mention it in "Crime and the Economy Don't Tell the Whole Story," ( Los Angeles Times, January 9, 2008) but that Depression experience was paralelled in the UK 50 years later, as some crime rates fell significantly during the bleak years of rising  unemployment when the Thatcher reforms first took effect.

One thing is reasonably sure for New Zealand. Our disgraceful crime rates are not attributable to 'Rogernomics'. The rocketing growth was noticeable in the early 1970s, shortly after publication of  "Crime in New Zealand". That book set the scene for three decades of reforms to focus our criminal justice system away from punishment, and on to therapy for the offender.

My conclusion, from all the reading and visiting and listening which was for me the greatest privilege of being an MP, was that unfounded belief in the State's power to cure criminal tendencies, and loss of focus on what the State clearly can do (punish to deter), explains more than any other factor.