Tim Lambert (deltoidblog AT gmail.com) is a computer scientist at the University of New South Wales.
I decided to check the claim that crime went up following gun control in England in 1920. I compared the homicide rates for England for the period 1911-1920 with the period 1921-1930 and found that the rate declined by 8%. This change could have been caused by demographic changes that followed from World War I, since this killed a significant number of young males (the group that murders most.)
To check for this, I looked at all the countries in [1] for which homicide data was available covering 1911-1930. A feature common to almost all these countries (Japan being the major exception) was a pronounced dip in the homicide rate in the years 1915-1918, presumably because large numbers of young males were serving in the war, so I excluded these years from the analysis, to get a better idea of the true homicide rate before 1920. We also want to establish if any changes are significant, so I did a t-test for each country.
Here are the results:
11-14 21-30 change Prob chance could
19-20 % cause this change
England 0.80 0.69 -12 0.003
Australia 0.89 0.67 -25 0.05
Germany 0.58 0.84 +41 0.001
Italy 7.85 10.2 +29 0.24
Japan 3.46 3.78 +9 0.14
Portugal 2.82 3.16 +12 0.17
Scotland 0.74 0.78 +5 0.55
USA 7.52 8.58 +14 0.006
(Note that some of the rates are for homicide offences and some for convictions, so you cannot compare rates across countries.)
We see that in most countries the homicide rate increased, with Germany and the US recording significant increases.
The only country beside England to record a significant decrease after 1920 was Australia. NSW (40% of Australia's population) introduced gun control in 1920. Victoria (30%) followed in 1921, and Queensland (15%) in 1927.
