https://www.nytimes.com/2018/05/12/world/australia/mass-shooting-osmington.html

But if it makes you feel better, Australia does indeed have a significantly lower murder rate. However, its worth noting a good chunk of US murders aren't done with guns. Some FBI stats I've seen show about 50% is not done with a gun. Statistically your odds are pretty high of being beaten or stabbed to death as opposed to shot.

There's been some this year... Margaret River shooting. 7 dead. Here's a link with several more.

https://www.watoday.com.au/national/western-australia/margaret-river-tragedy-australia-sees-its-worst-mass-shooting-since-port-arthur-in-1996-20180511-p4zere.html

http://www.foxnews.com/world/2018/05/14/grandfather-is-suspect-in-australia-mass-shooting-official.html

The reason I keep ignoring the areas with high homicide rates and low population is because they don't really matter very much to the overall point. If the point is "many Americans do not have to worry about homicides", I don't see why a small portion of the population having to worry about homicides really matter. What matters most are the areas where people don't have to worry about homicides; and the areas with high population.

Maybe you are misunderstanding me. I am not arguing about cause here. I am not saying that the reason the center of the USA doesn't have a lot of homicides is because of its low population density. While there is certainly a trend between population density and homicide rate, I am not discussing these causal factors here. I am merely stating that the low homicide rate in the middle of the country is not representative of a low homicide rate for an amount of individuals proportional to the area of the center of the country.

Also, I disagree with the idea of using the population density of a whole state as a means of determining this correlation. States vary greatly in population density. Look at Texas for example. Does it make sense to say that someone in Dallas is safer because of what is happening in Sanderson (on the Western side of the state)? Not really. Those two points are almost 500mi apart. The vast variety in population density across the state would greatly skew those numbers.

Finally, are you trying to make the argument that cities don't tend to have higher crime rates? Am I misunderstanding that?

Primary or not, gangs are a big chunk of the total murders. It also depends on the city, some areas have bigger gang problems than others.

Chicago for example has a huge gang problem and that single city accounts for a fairly large percentage of overall murders in the country.

I wanted to do some math before I responded to this, because it is an interesting question: How much of a difference does 0.22/100k homicides make.

So, before I get started here are the numbers I used:

Homicide Rate of Canada: 1.68/100k
Homicide Rate of Massachusetts: 1.9/100k
Population of Canada: 36million
Average life span of a Canadian: 82 years
Number of people the average person knows ( http://www.stat.columbia.edu/~gelman/research/published/DiPreteetal.pdf ): 600

With those variables, I calculated the number of homicides in Canada using the two different rates: 605 and 684
Because this is a yearly rate, I multiplied this by the average lifespan to find how many people are killed over an individuals lifespan: 49610 and 56088
Then I multiplied this by how many people the average person knows: 29.8million and 33.7million (I think that this is right, but I've started doubting myself. Basically, there are about 60,000 "pools" of 600. About 50,000 have one individual who has been victimized, leading us to the above numbers. You could likely calculate a more in depth probability here but I think this is right if we assume a perfectly even distribution).

So, the difference is that an extra 4ish million people in Canada would know someone who had been the victim of homicide throughout their lifetime. In percentages, that would be 83% vs 94%, so an extra 11%.

Obviously that math makes a lot of assumptions about the simplicity of the data, so take it as more of a curio than hard facts, but I found it fairly interesting.

All that said, I'd wager that, while knowing someone who had been the victim of homicide would certainly be a large influence on how much someone notices a changing homicide rate, factors not necessarily directly tied to the rate like media coverage, media engagement and general exposure to crime would certainly play quite a big part too, so overall, I don't really think it is possible to say how much of a difference 0.22/100k would actually make.

Oh in that case, to get a good idea, I added up the populations of counties with homicide rates <2.0 using data found in the following link. https://wonder.cdc.gov/controller/datarequest/D76 . 

I downloaded the data, converted it from a text file into an excel file, and then wrote a simple SUMIF statement that said, "if Crude Homicide Rate < 2.0; sum up population." 

Doing that I got ~ 41 million Americans live in counties with homicide rates that average less than 2.0  over a period of 17 years (1999 - 2016.) This number would be larger if we recognize that the homicide rates have declined by about 20% country-wide since the late 90's, so if I could select only for 2016 data (there isn't enough data for 2016 alone), we would probably expect that number to be higher, by how much? I can't say.  

If we increase the rate to <2.5, we get ~ 64 million Americans. If <3, we get ~86 million Americans, if <3.5 ~ 117 million Americans, if <4.0 ~ 136 million Americans, if < 5.0 ~169 million Americans, if < 10 we get ~ 289 million Americans, if < 15 we get ~ 311 million Americans, and if < 25 we get 322 million Americans, if < 45 we get 325 million Americans. 

So about 12.5% of Americans live in counties with homicide rates less than 2.0 per 100,000. 

About  42% of Americans live in counties with homicide rates less than 5.0 per 100,000 homicides. (This should be ~50% because the American homicide rate was 4.9 per 100,000 in 2016, but remember that this is data for 17 years. The homicide rate in 1999 was 5.7. ) 

About 90% of Americans live in counties with homicide rates less than 10.0 per 100,000 people. While that sounds sort of bad, in the 1980's 10.0 per 100,000 people was the average nationwide. So a lot of progress has been made. The question is: how can we continue? 

Favorably comparing yourself against 3rd world countries sure is a sign of great strength and accomplishment.

If you compare USA to Mexico and Brazil of course it is much safer and much less gun violence lol. I coulda told you that without your research, thank you. 

The point is USA is much more dangerous in terms of gun violence than comparable, developed, 1st world countries such as Germany and Japan.

Is there gun violence in Germany Japan? May be a few times a year. In the USA, gun violence is every day. 

Do I fear of getting shot in Germany and Japan? Nope. In USA, yes. 

Alright, well mathed. That is pretty much the type of result that I expected. I feel like using that data request form you should be able to limit the data to only 2015, but after a few minutes of tinkering I couldn't quite figure it out and I think what you've provided is adequate to answer the question I wanted answered.

By the way, I did make another reply to you if you missed it. It was just a random curio, so theres no need to respond to it but I just wanted to let you know in case it was missed.

I stand corrected!
Still a good track record since gun control was brought in during the 90's thanks to Johnny Howard.
So it still worked. - How many years before the next one though?