There are two reasons Moore may be reluctant to talk numbers rather than rates or probabilities.

First, the other countries he compares have a smaller (sometimes a much smaller) population) than the US. The US has a population of about 285 million: Canada about 31 million, Australia 19 million, Germany 81 million. For example, in 2000-2001 Australia reported 381 homicides among its 19 million people. Source. If we were to apply its rate to the US population, that'd work out to 5,715 murders. That'd still be lower than the US actual, but since the Aussie rates are skyrocketing, they may catch up soon.

(The above source puts a peculiar spin on the increase, by the way. Although reporting a massive 20% jump in murders, and a 100% increase in intra-family killings, it is headlined that "In 2001-2002 there was a 25 per cent decrease in the use of firearms to commit homicide." Great: what it really means is that gun homicides fell slightly, but were more than offset by increases in all other means of murder. The above figure of 381 is, by the way, total homicides rather than gun homicides.... it's of no benefit to get killed with a knife or club, either.)

Second, the minute you get into rates it starts to become apparent that gun homicides are a very low probability event wherever you are. Phrased otherwise, the odds are enormously in favor of you or I dying of heart attack, stroke, or cancer, no matter what country we live in.

Here's the 1999 Center for Disease Control statistics (which is what Bowling uses) for the US:

Deaths from heart attack &

other cardiovascular failure: ........ 892,558

Deaths from cancer: ................... 549,838

Deaths from flu & pneumonia: .... 67,730

Deaths from kidney failure: ......... 35,525

Deaths from diabetes: .................. 68,399

Deaths from alcohol ingestion .....19,171

Deaths from gun homicides: ....... 10,217 (includes self-defense and police killings of criminals)

Deaths from gun accidents .............. 824 (not a typo)

Of the 2.39 million US deaths that year, gun homicides comprised 0.4%.And of course that probability can be sizeably reduced by a few lifestyle changes. E.g., don't deal crack. Gun accidents comprised three one-hundreds of one percent.

But you can't make much of a documentary around a theme of "The US has an enormous problem -- almost one-sixth as likely to kill you as the flu, and one-ninetieth as likely as heart failure."