Today Thu Fri
It is forecast to be Chance of Rain at 11:00 PM EDT on August 20, 2014
Chance of Rain
It is forecast to be Thunderstorm at 11:00 PM EDT on August 21, 2014
It is forecast to be Mostly Cloudy at 11:00 PM EDT on August 22, 2014
Mostly Cloudy



Metrocide: Location, Location, Location

Is Toronto a dangerous city? And is it getting worse? This week for Metrocide, Torontoist is examining a sea of homicide data and trying to come up with conclusions based not in fear or fantasy but fact.
Photo of Toronto from the CN Tower in 1981 by retroman from the Torontoist Flickr Pool.
Though it’s difficult to ethically explain away, homicides tend to affect people less the further away they happen, on both on a global scale and a local one. That’s why, perhaps, downtown homicides undoubtedly receive a disproportionate amount of attention: not only do a lot of people live downtown, a lot of people go downtown. The result of all that attention? The perception that, as one commenter put it yesterday, “Toronto suffers murders on Yonge St, in the club district, and on other downtown streets on a regular basis”—even though it measurably doesn’t.
Today, we’re examining homicides by narrowing our focus onto downtown and central Toronto, in an attempt to see if either of those areas are any more or less dangerous than the rest of the city. Tomorrow, we’ll go the other way, and compare Toronto to other cities to see where we stand.

Notes on Statistics

If you haven’t yet, be sure to read yesterday’s note on statistics, which explains the different geographic areas of Toronto that Toronto Police and Statistics Canada data represent. Today’s statistics are almost entirely drawn from Toronto Police data—only their data break down homicides into smaller fixed geographical boundaries than just “Toronto”—so heed the organization’s own warnings that “statistics…count all offences reported to the police….a single incident reported to police may generate more than one offence,” and that individual year-to-year statistics are less important than “the general trends and magnitude of change.”
Only the past ten years of data are publicly available on the Toronto Police website, so a larger picture is substantially harder to see than it was when we had twenty-six years of StatsCan data to work with yesterday. We’re also missing some data from 1998: the police’s annual statistical report for that year is, of yet, only available online as a PDF with a table of contents but no actual data. They haven’t gotten around to fixing it in time for us to publish Metrocide. (We did ask last week, but presumably the police have slightly more important things to do.) Please note that if there are no numbers charted for 1998, it is because those numbers are not known, not because they don’t exist. We’ll do what we can with the statistics we have.
The focus of today’s article is two areas we’re calling central and downtown Toronto. The Toronto Police define central Toronto as the area that extends from the Humber River in the west to Victoria Park Avenue in the east. From the Humber east to about Bayview Avenue, the area’s northernmost point is Lawrence Avenue; once Lawrence hits Bayview, the border of central Toronto winds down the Don River, until the Don River hits Eglinton Avenue East. (On the Toronto Police map, it’s the beige-coloured area). Those areas are served by nine police divisions: 11, 12, 13, 14, 51, 52, 53, 54, and 55.
Though what does or doesn’t count as downtown Toronto is open to interpretation, the geography of Toronto Police divisions all but force those boundaries to be those of 51 and 52 Division if we want to look at the relevant statistics. 51 Division covers the area from the lake north to Bloor Street East, and from Yonge Street east to the Don Valley Parkway, and 52 Division covers the area from the lake north to Bloor Street West, and from Spadina Avenue east to Yonge Street.
So, what do the statistics say?
The number of homicides that occur in central Toronto, at least over the past ten years, does seem to be slowly rising, though it is just as inconsistent on a year-to-year basis as Toronto’s overall number of homicides since 1981. Last year, for instance, there was one fewer homicide in central Toronto than there was in 2000, and one more than there was in 1997.
Overall, however, the percentage of Toronto homicides that take place in central Toronto seems to be declining, weighed against both Statistics Canada’s and the Toronto Police’s definitions of Toronto. The share of homicides that occur centrally is unpredictable and inconsistent, too: note the jump in 2006 before the fall in 2007.
The central homicide rate seems to hold steady with the city’s overall homicide rate as measured by Toronto Police, though it is not a perfect correlation: it seems to have originally been higher than the overall homicide rate, and to have recently dropped.
Also of note: both central Toronto’s homicide rate and Toronto’s homicide rate (as both measured by the Toronto Police) are higher than Metropolitan Toronto’s rate as measured by Statistics Canada, which demonstrates that, if they are considered together, Toronto’s satellite cities and towns not included in Toronto Police data (Mississauga, Oakville, Brampton, Newmarket, Caledon, Markham, and so on) have a lower homicide rate than Toronto proper does.
While the number of homicides in central Toronto seems to be increasing, there seems to be no pattern whatsoever to homicides in downtown Toronto (as we’ve defined it) over the past ten years—there have been as few as six homicides (in 2002) and as many as 17 (in 2000). In 2007, there were as many homicides downtown as there were in 1999, and less than there were in 1997, 2000, 2001, 2003, 2005, and 2006.
Since Yonge Street divides 51 and 52 Division, and since the two divisions are roughly the same size, it’s also worth noting that more homicides that occur downtown occur east of Yonge than west of it. That means that 52 Division (coloured in green above)—which includes the Entertainment District, the Financial District, as well as pretty much all of the city’s prime tourist attractions—has consistently fewer homicides than the significantly more residential area patrolled by 51 Division (coloured in blue).
The perception that homicides often occur downtown seems to be wrong, according to the data. On average over the nine years charted above, someone is killed in 52 Division every 107 days and in 51 Division every 47 days. So far this year, there has been one homicide in 51 Division, and have been none at all in 52 Division—both slightly less than there were by July 23 in 2006 and 2007.
Even though the number of downtown homicides is inconsistent, it seems that the percentage of homicides that occur in that area compared to the rest of the city (or to central Toronto) seems to be declining from a peak reached seven years ago. In other words: regardless of whether or not you think that downtown Toronto is deadly, it seems to be consistently getting less and less so compared to the rest of the city.
Unfortunately, it’s difficult to find a discernible homicide rate for downtown Toronto—the Toronto Police include a central population in their statistical reports each year (which can then be used to determine the homicide rate per 100,000 people), but they do not currently include year-to-year populations of individual divisions. Nonetheless, it seems reasonable to assume that the population in downtown Toronto has gone up over the past ten years, and will continue to do so into the future. Even if a rate were available, it may not be a good idea to treat it as gospel: all of the homicide rates we’ve included here are determined by the population (i.e. the people who live in a given area), and not the amount of people who spend a significant chunk of time working or hanging out there—which, for downtown Toronto, would almost certainly make the population significantly larger and the homicide rate significantly smaller.


All Toronto Police statistics from 2007 and earlier are from their year-end statistical reports. Year-to-date crime statistics for 2006, 2007, and 2008 are from the statistics area of their website.
Metropolitan Toronto homicides from 1981 to 2007 are from Statistics Canada. (Table 253-0004—Homicide survey, number and rates [per 100,000 population] of homicide victims, by census metropolitan area (CMA), annual, CANSIM [database].)


  • ysdn

    T.O. is SAFE.
    if youre scared, move to philly/dc.

  • matty

    Man this feature is kick ass. Nice work again David Topping. I think you’ve made your case though!

  • Careygrrl

    Dudes, your charts are very nice, but 29 homicides a year is too many, even if it is down from 36. Period.

  • David Newland

    This is good work Topping, and I commend you for it. But you need to go into details about the TYPES of homicides and their locations, not just vaguely, but closely.
    For example, suppose there used to be 50 homicides yearly, committed with knives, behind closed doors, and the victims were all drug dealers; also suppose that today, there are 30 homicides yearly, out in the street, with guns, and the victims are frequently people all minding their own business.
    In that scenario, there’s a quantitative downturn, but qualitatively, I have much more reason to be afraid.
    In my “safe” neighbourhood at least 2 people were killed in shootings in broad daylight over the past year. In one case, the victim seems to have been the target (though it’s a miracle bystanders weren’t killed); in the other, the victim was an innocent man.
    You need to reassure us that this situation is not becoming more common before these stats go any distance toward making the city less scary.
    And Careygrrl is right. 29 is too many, period.

  • David Topping

    Of course any number of homicides above 0 is too many.
    Unfortunately, David, the amount of detail you’re asking for—a breakdown of how many victims of homicides are drug dealers, for instance, or how many victims are killed “behind closed doors” versus on the street, or how many people are killed on this street versus that street—is, as far as I know, data that is not compiled and that would thus require my looking at each individual homicide that has occurred over however many years would be necessary to determine some kind of pattern (so, probably more than ten). It would be fantastic if someone were able to do that work, but I can tell you that the process of compiling, examining, and presenting the data for the Metrocide posts has already taken a solid week of work, and that what you’re asking for would probably take someone months.
    If it’s something you’re interested in, though, I suggest that you get in touch with the Toronto Police’s Access and Privacy Section and look at homicide records for your neighbourhood. Their contact information is here.

  • David Newland

    Fair enough. Your effort truly has been noted.
    I should have made my point differently; rather than suggesting that _you_ need to do that additional work, I should merely have said _someone_ needs to do it before the point can fairly be made that Toronto is relatively (let alone absolutely) safe.

  • uskyscraper

    Well put, David N. The best way to try and discern type without the real data is perhaps to look at geographic areas that do not include large residential populations – i.e. the downtown office/tourist/entertainment cores of cities. Based on the chart above one might deduce that 2-7 homicides per year are happening in this zone in Toronto, and by nature of the location these tend to be murders in public or semi-public spaces. I believe these numbers to be comparable or higher than in some other cities, regardless of the overall citywide murder rates. For example, murder-happy Houston has not had a single homicide in the area around their downtown core yet this year [], while Toronto has had five in a similar zone (Bathurst to DVP, south of Bloor), four of which were on the street [].
    My original line about “Toronto suffers murders on Yonge St, in the club district, and on other downtown streets on a regular basis” is based on a n admittedly very high expectation, and just a few murders a year in public is still what I would consider both “a regular basis” and unacceptable for law-abiding citizens, visitors and tourists.

  • rek

    I’ve never been to Houston or New York or London or Berlin, but I know from my experience in Chicago and Seoul and Toronto that ‘downtowns’ are not so easily compared. I was going to reply with this in the last Metrocide but figured it would be more suitable here given the topic.
    All sorts of factors can contribute to the ‘downtown’ homicide rate a city has, and make for very different places despite being called ‘downtown’. Psychogeographic barriers, economic barriers (road tolls), property value, ease of access (transportation), concentration of attractions and competition from other parts of the city, mix of high- and low-rent attractions, comparative residential and non-residential usage, number of public gathering spaces, police visibility…
    Murder-happy Houston might not have had a murder downtown, but does anyone live there, is it the kind of downtown that closes up at 6pm, is there any reason for people to gather in large numbers there when they aren’t working? Chicago’s Loop has had a single murder in the past 365 days, but it’s also a racially and economically (rich, white) homogenous zone. Include neighbourhoods bordering the Loop, where regular people live and work and go clubbing, and the number of murders jumps to 36 in the same period.
    In 2007 nobody was killed on Bay Street, but go a few blocks east or west and the bodies begin.

  • McKingford

    there are 30 homicides yearly, out in the street, with guns, and the victims are frequently people all minding their own business.
    Please – some perspective.
    Jane Creba and John O’Keefe. Those are the two cases that shocked Toronto because they were innocent bystanders minding their own business. And they represent 2 murders in 2 years.
    The overwhelming majority of murders continue to involve people who were not simply minding their own business, even if we collectively freaked at those other two cases.

  • TokyoTuds

    Although any murders are too many, it is naive to say so, as human nature will insure future murders. Crimes of passion will happen, and people will get murdered with guns, knives, a candlestick, whatever …..
    The reason I favour gun bans is that in a crime of passion, the victim is more likely to survive and fully recover physically from a candlestick attack than a gun attack.
    My final point however, is that In Canada (and Toronto, I’m sure) you are more likely to kill yourself, or die in a fall, than to be murdered.
    2004 Rate per 100,00 for Suicide 10.8
    2004 Rate per 100,00 for Falls 5.4
    2004 Rate per 100,00 for Murder 1.7

  • rek

    Homicides by accused-victim relationship, 2006
    ## Victim killed by
    91 Casual acquaintance
    75 Stranger
    62 Current Spouse (includes common-law)
    54 Criminal relationship
    42 Close friend or neighbour
    41 Other family
    31 Parent
    16 Ex-spouse
    15 Child
    15 Boyfriend/girlfriend (current or former)
    07 Business relationship – Legal
    03 Unknown
    605 Total homicides, Canada
    I imagine these numbers also reflect the average accused-victim relationships in Toronto homicides (Toronto CMA 2006 homicide rate was 0.02 lower than the national average), and so long as they aren’t different by a significant amount they clearly show you’re more likely to be killed by someone you know than someone you don’t (530 of 602 known relationships, or 88%).

  • David Newland

    McKingford, quoting my illustration without the words “suppose that” is misleading. I’m not suggesting there were 30 innocent people murdered, but that if there had been, it would surely be a qualitative difference from fifty druglords being murdered.
    When Hou Chang Mao was gunned down in broad daylight on a busy street in my neighbourhood, my sympathy with the Polyanna view ran out.

  • McKingford

    David, sorry – I didn’t mean to change the context of what you said.
    My simple point is that the type of murder that truly scares us is *random* stranger-on-stranger murder. And that type of murder is exceedingly rare; the cases of Jane Creba, John O’Keefe and Hou Chang Mao are the exceptions that prove the rule.
    I appreciate the stats produced by t.rek, which showed 75 murders by strangers in 2006. I have to think that *random* murders by strangers (a la Jane Creba) remain a tiny subset of those. The majority of murders by strangers still involve non-random altercations (eg. drunken late night argument in the club district).
    The fact remains that, despite the media frenzy to the contrary, it is still exceedingly safe to be an innocent bystander.