What has misery got to do with murder?

A model of what contributes to and also what reduces murder

Image: ‘Misery’, AKA Tanja Thompson, acclaimmag.com

This is a longer article about the socioeconomic factors contributing to New Zealand’s murder rate. Introducing the ‘misery index’ and evolutionary theories of competition between young males.

Why was 1959 such a special year in New Zealand?  Apart from the opening of the Auckland Harbour Bridge on 20 May, it might seem that 1959 was quite unexceptional.  There was a happy mix of low inflation (around 0.7%) and low unemployment (about 1%).

Britain bought nearly 50% of our exports at this time. Harold MacMillan described this to Charles De Gaulle a few years later (1962); New Zealand was “an English farm in the Pacific.” 

There was serious racial inequality, evidenced by the difference in life expectancy between Māori and non-Māori in 1961. Māori males lived 10.2 years less than non-Māori, and Māori women missed out on 13.1 years. 

The remarkable but unnoticed thing about 1959 is that there were no murders.  New Zealand Police has a data series going back to 1926, 1959 is the only zero-murder year. 

Until the 1970s the New Zealand murder rate was typically below 1 per 100,000 people per year (except for 3 years in the early 1930s and 2 years in the mid-40s when it went above 1.0).  Then it moved up, through 1.5 in the early 80s, then around 2 from 1985 to 1992, before beginning a trend down towards 1 (I am excepting the 2019 surge resulting from the Christchurch mosque murders).  Is there any explanation for these long-term trends?

There are different types of murderers.  There are different paths to their crime and different victims.  The most common murder is committed by a young, impulsive and thought-disordered male.  Impulsiveness is the product of drastic discounting of future consequences.  The thought disorder is the misreading of a situation as representing an existential threat.

The rate of murder varies incredibly across different times and different societies.  (There is some indication that it is the impulsive or reactive type of murder that is the type that varies so much – but that’s another story.) Over centuries in Europe, the murder rate has dropped.  One estimate for London in 1278 is 15 homicides per 100,000 souls. Now most European countries have murder rates below 1.  The US has significantly higher murder rates than Europe – at present around 6 / 100,000.  Within the US the murder rate is significantly higher in ‘southern’ than ‘northern’ states – from a high in 2020 of 15.8 / 100,000 in Louisiana to 0.9 in New Hampshire.

Why does the rate of murder vary so much from era to era and across societies?

Here I look at New Zealand’s history to see if there are clues to changes in the rate of murder over the years.  What social trends fit with the trends we see in New Zealand’s changing rate of murder from 1926?  While New Zealand’s murder rate is low and not especially variable compared to other countries, we have been through some big changes in the past 100 years (the period of reliable murder statistics).  A warning – I am going to talk about correlations and use graphs to illustrate the possible role of socioeconomic factors.

The evolutionary view is that young men competing for mating opportunities make up the majority of both perpetrators and victims of homicide. The trend line for the proportion of young men aged 15-29 does show some match to the murder trend – fewer young males go with a lower murder rate, and vice-versa (murder rate is presented as ‘per million’ to make for a tidier graph).  The correlation between the proportion of young men and the murder rate over the years 1926 to 2001  is a small .22 (the convention about interpreting such correlations is that if it’s over .20 it's small but could be meaningful).

Taking this evolutionary view further, if women are delaying marriage, as seems to be a trend since the late 1960s, then it becomes more competitive for men, increasing the risk of more lethal aggression.  The correlation between murder and women’s median age at marriage over the period 1961-2018 is another small but interesting .33 – women marrying later goes with more murder, at least until the 1990s.

Some economists make a connection between well-being and the economic factors of inflation and unemployment.  The combination of inflation and unemployment is known as the ‘misery index’ – when both go up, well-being (as measured by happiness ratings) goes down. 

In New Zealand, inflation (the consumer price index or CPI) took off in 1974 and stayed high (excepting 1983 and 1984) until 1987.  More inflation correlates with more murder – at a small correlation of .30 (for the years 1960-2018).

Unemployment rose in the early 1930s (surprise, surprise – this was the years of the Great Depression), as did the murder rate.  For the first time in decades, unemployment went up to and above 10% over the period 1991-1993 (the period of New Zealand’s massive socioeconomic restructuring).  The correlation of unemployment with the murder rate is ‘moderate’ at .53. However, unemployment peaked years after peak murder; murder rates were around 2 / 100,000 from 1985-1992.

Combining inflation and unemployment (creating a ‘misery index’ from their respective ‘z-scores’, sorry I didn’t warn you about that – this transformation is just a way to ensure we are getting a balanced combination of the two measures), we see a tighter relationship with murder rates.  The graph shows a closer sync through the years between the misery index and murder – both rise in the 30s. drop through the post-war years, climb together in the late 70s, and then slowly drop away after the 80s. This is summarised statistically by the moderate .54 correlation between the misery index and the murder rate.

Who is hurt hardest by rising inflation and unemployment?  It is those at the bottom of the social hierarchy.  Inflation (‘the thief in your wallet’) means life’s essentials are less affordable. If you are poor you don’t have other assets you can convert to cash to help you through the current hardship.  Getting work is not always a way out. Almost by definition those at the bottom have fewer qualifications and although cheaper to hire they are also more replaceable.  New Zealand research has found that work is more precarious for the poor.

Are there any other indications of this view that squeezing those near the bottom can result in more violence? In New Zealand and much of the world the last century saw radical increases in life expectancy.  But while the overall trend is positive there have been bumps, most notably the reversal of this trend for Māori in the 1980s.  The gap between Māori and ‘non-Māori’ life expectancy (LE) had been narrowing spectacularly until the 60s.  This relative improvement slowed and then reversed in the 80s.  Māori LE was now getting worse relative to non-Māori.

The chart of year-on-year relativity in life expectancy between Māori and non-Māori shows the gap between Māori and non-Māori widening over the 1986-1996 decade – indicating that things were getting relatively worse for Māori. There is a corresponding rise in murder rates (1985-1992).  There is a moderate (-.45) correlation (a reverse correlation to be technical) between these measures – relative decreases in Māori life expectancy go with increases in murder rates (over the period 1951-2011).  Because life expectancy results are not available every year, this correlation rests on only 13 data points, caution is necessary.

This correlation between the rise in murder and the decrease in Māori well-being is an indication of how murder can be considered the result of serious pressure on those lower down the social hierarchy. (LE as a measure of well-being? You can’t enjoy life if you’re dead!) 

As life gets tougher and more competitive, young males are increasingly sensitive to the threat of being pushed further down the pecking order. For those with limited personal and social resources (intelligence, impulse control, qualifications, supportive social networks) finding themselves accelerating down with little prospect of status and marriage can mean taking an increasingly short-term view, more inclined towards desperate action, including violence. They are exquisitely sensitive (often potentiated by alcohol or drugs), so even trivial threats to self-respect can trigger violence.

Marvin Wolfgang, in 1958, describes the most common motive for murder: “an altercation of relatively trivial origin, insult, curse, jostling etc.” Trivial yes, but seen as an existential threat by vulnerable, young, unmarried males.

I hear some of you – this calls for a multivariate analysis Stewart. Well, maybe I don’t hear that! Meanwhile, do you have a theory about what contributes to higher (or lower) murder rates? I’d love to hear your thoughts.