Since the terrorist attacks of Sept 11, 2001, economists and others have crunched a lot of numbers trying to identify the root causes of mass violence.

Their work has largely dispensed with the theory - advanced by George W. Bush in 2002 and seconded by various members of the current administration - that poverty is a major cause of terrorism.

From a 2009 Rand study prepared for the Defence Department: "Terrorists are not particularly impoverished, uneducated, or afflicted by mental disease. Demographically, their most important characteristic is normalcy (within their environment). Terrorist leaders actually tend to come from relatively privileged backgrounds."

Economic growth seems like it might be a more promising variable - when it slows down, there are lots of disappointed people with lots of time on their hands - but so far, the data on that seems pretty inconclusive, or at least confusing.

Unemployment and underemployment, especially among the young, are more direct measures of the same phenomenon, and there is some evidence that those factors are linked to violence. Sometimes the link goes in the opposite direction from what you might expect, though: A study of domestic violence and unemployment rates in the United Kingdom found that higher male unemployment was correlated with reduced violence. Also, the labour market data available in developing countries - where the vast majority of terrorist activity has occurred in recent years - is not reliable enough to justify strong conclusions.

The root-cause explanation with the strongest evidence to back it up may be simple demographics. The theory here is that young men have a far greater propensity to commit violent acts than any other demographic group, so when there are more of them around - when a country experiences a "youth bulge" - there's more likely to be trouble.

Sure enough, multiple empirical studies have shown that, as Norwegian political scientist Hendrik Urdal put it in 2011, "youth bulges are associated with an increased risk of political violence".

The bulk of the research cited above has focused on developing countries. As already noted, that's where most of the action has been. But what about the United States, which has been beset in recent months by an alarming series of killings - some widely labelled as terrorist and some not, but all with the aim of spreading fear - committed almost exclusively by young men?

Economic growth in the US has slowed markedly since 2000. The percentage of young men with jobs is well below what it was in the 1990s or even 2000s: Among those aged 20 to 24, a lot of that decline has to do with more people attending college and, in recent years, sticking around for grad school. But that's less of a factor for the 25- to-34 cohort, and it surely does not explain all of the decline for the 20-to-24-year-olds.

In general, the labour market has become less welcoming to men in recent years. As a result, there are more young men with time on their hands than there used to be - especially since the US has been going through its own youth bulge: So that should explain it: youth bulge + fewer jobs for young men = murder and mayhem.

Except that this does not quite add up for the US. For one thing, the youth bulges Urdal and others have studied in developing countries have been much, much bigger than what the US has been experiencing. For another, broad measures of violence in the US are declining, not increasing. The gun homicide rate has been falling for years, as have arrests for violent crimes.

The peak age for murder arrests in the US was 19 in 2010, according to the Justice Department. The number of 19-year-olds arrested for murder that year was well below the number in 2000 - even though the youth bulge was near its peak in 2010, meaning there were a lot more 19-year-olds in the US than there had been a decade before.

From the Justice Department report: "While the period from 1990 to 2000 saw large declines across all age groups, the overall decline in the murder arrest rate between 2000 and 2010 was primarily the result of continuing declines in arrests for older juveniles and young adults (ie, persons between 17 and 29 years old)."

Maybe that is because young people have become more adept at getting away with murder. I doubt that, though. I also doubt that we're ever going to get a convincing root-cause explanation for the disturbing paroxysm of high-profile violence that the US has been going through.

BLOOMBERG VIEW