Mundt, Gun Control, Canada and the US

Canada. Gun law in 78.

Homicide rate (per 100,000 population)
74-78 2.7 3.1 2.9 3.0 2.8 average 2.9
79-83 2.7 2.5 2.6 2.7 2.7 average 2.6

(a t test on the statistical significance of the difference of the means gives p=.01)

Thomas Grant Edwards said:

From “Gun Control and Rates of Firearm Violence
in Canada and the United States” by R.J. Mundt, in Canadian
Journal of Criminology
, Jan. 1990, p. 137:

“The mean rate [of homicide] for Canada from 1974-1978 was 2.7,
compared to a post-1978 rate (through 1988) of 2.6. One could
admit the possibility that this decline resulted from the 1977
legislation, except that the mean rates for the United States
in the same periods dropped from 9.2 to 8.9, and almost identical
drop.”

Wait a moment. I got 2.9 as the mean homicide rate Canada 74-78, not
2.7. Something is wrong here.

Let me have a look at Mundt’s paper…..

Oh my goodness. Mundt has messed up badly.

Let me try to explain what has happened. There are two ways to
count homicides. You can look at death certificates or police
records. The numbers are a little different, but it doesn’t matter as
long as you don’t mix them up.

I got the death certificate counts from the World Health Organization
Statistical Yearbook. I got the police record counts from “The Size
of the Crime Problem in Australia” (which got them from Statistics
Canada 88). Mundt presents his data in the form of a graph. I extracted
the numbers from this. In the graph, a 0.1 homicide rate corresponds
to less than a millimetre, so these numbers are +/-0.1.

                    74  75  76  77  78  79  80  81  82  83  84  85  86  87
death certificates 2.5 2.7 2.4     2.5 2.5 2.1 2.3     2.4 2.3 2.1 2.1 2.2
             Mundt 2.5 2.8 2.4 2.6 2.5 2.5 2.1 2.4 2.7 2.7 2.7 2.8 2.2 2.5
    police reports 2.7 3.1 2.9 3.0 2.8 2.7 2.5 2.6 2.7 2.7 2.6 2.7 2.2 2.5

Can you see what has happened here? Mundt has used the lower death
certificate numbers for 74-81 and the higher police report numbers for
82-87. This has the interesting effect of canceling out the decrease
evident if either death certificate numbers or police report numbers
are used.

How did Mundt go wrong? He states that the source of his data was
Scarff 83 and Statistics Canada 88. From reading the paper you can
discover that “Scarff 83″ covered 74-81, so this is the source of his
death certificate numbers, and of course “Statistics Canada 88″ is the
source of my police report numbers as well. Mundt’s error is
inexcusable, since the “Statistics Canada” reference contained data
for 74-81 and he failed to compare these numbers with the “Scarff 83″
numbers.

It is also interesting to compare firearm homicide rates and
non-firearm homicide rates. After all, if both firearm homicides and
non-firearm homicides decreased by the same percentage, it would be
unreasonable to conclude that the gun law had any effect.

                                                                  ave   ave
          74  75  76  77  78  79  80  81  82  83  84  85  86  87 74-78 79-87
firearms 1.4 1.4 1.2 1.2 1.1 0.9 0.9 0.8 1.0 0.9 0.9 0.8 0.7 0.8  1.3   0.9
other    1.3 1.7 1.7 1.8 1.7 1.8 1.6 1.8 1.7 1.8 1.7 1.9 1.5 1.7  1.6   1.7
methods

t-tests show that difference in the means for firearms is significant
(p=0.00003) while the slight increase for other methods is not
(p=0.39).

Of course, here is the big kicker in Canada: Let’s say you do
believe that the pre-1977 murder rates are more than the
post-1977 murder rates enough to make it worth your while.

“…there are now 1,275,000 more firearms in private ownership
in Canada than when the Act was passed — about 11,960,000, or
46,000 per 100,000 population.”

Mundt has only counted gun purchases, he has not allowed for guns
being sold/junked, so I don’t see how he can draw this conclusion.
In any case, number of guns per 100,000 population is an almost useless
measure of gun availability. By this measure guns are equally
available in the case where each of ten people has one gun and the
case where one has ten and the other nine none. To put it another
way, the gun purchases could have been mostly by people who already
had guns.

“The stock of restricted weapons (almost all handguns) increased
from an estimated 651,000 in 1976 to 923,000 in 1988 (based on the
total number of restricted weapons registered with and reported by
the RCMP.”

There doesn’t seem to have been any change in the handgun homicide rate.

So whether or not you believe murder went up, guns certainly did.

The relevant statistic would seem to me to be the percentage of the
population with guns, and we have no idea whether that went up or
down.

“When compared with the United States, trends in Canada over the
past ten years in various types of violent crime, suicide, and
accidental death show no dramatic results, and few suggestions
of perceptible effects of the 1977 Canadian gun control
legislation. This is scarcely surprising, except that expectations
were high among the policy formulators, and some evaluations
perhaps tried too hard to give them satisfaction. The decrease
in use of firearms in robbery appears to be the only change
that stands out over time or in comparison with parallel trends
in the United States.”

And I’m going to have to take exception to this as well. Lester
(Psychological Reports 72(3 Pt 1) 787-790) has shown that firearm
suicides were increasing before C-51 and decreasing afterward, while
other suicide methods did not have significant trends.

When discussing accidental deaths Mundt says “Death rates from
firearms accidents … have been in long-term decline in both
countries (Figure 6), with the American rate remaining from two to
three times greater.”

Looking at figure 6, we see that both rates decrease from 74 to 78,
with the US rate being twice the Canadian rate both in 74 and 78.
After 78 however, the US rate declines less rapidly, while the
Canadian rate declines dramatically (by 75%!), with the result that in
the last four years of the data set (83-86), the US rate is SIX TO
SEVEN time greater. I’m puzzled as to what made Mundt think that it
was only two to three times greater.

Conclusion: C-51 may have caused significant changes in the homicide
rate, suicide rate, and accidental death by firearms rate in Canada.