All around the world, cities’ high density areas get a worse rep than they deserve crime-wise. When a newspaper decides to publish a story about the most dangerous places in town, high density mixed used areas tend to be highlighted and unfairly shamed time and time again.
Is Downtown Toronto, for example, a scary and violent place? Mainstream media reports sure depict it that way, but you should look at the data more closely before drawing any conclusions. While the graphs, ranks, and maps that news outlets publish are often quite relevant, the assessment of their significance is rarely taken seriously.
Look up a map of Toronto’s crime distribution and you will find a perfect example of this trend. The maps themselves are often valuable and well made, but their interpretation (or lack thereof) is rarely satisfactory.
The map below, published by the CBC, shows the number of assaults per 10,000 residents in the City of Toronto’s various neighbourhoods:
What do you see? Downtown Toronto, especially the area encompassing the Financial/Hospital/and Provincial Government Districts, is by far the most dangerous place in town as far as assaults are concerned. But is this an appropriate and accurate interpretation of these results? (Spoiler: It is not) The CBC dilligently posted a “methodology and limitations” section attached to their virtual map application, but there is nothing in there that suggests that the map may be misleading in this regard.
What the CBC should have pointed out (among other things), but did not, was that adjusting for residential density only gets you so far. The assaults reported to Toronto police were not necessarily committed by or against members of one of these neighbourhoods. What this means is that, assuming crime is spread homogeneously, after adjusting for residential density you will predictably end up with a map where the neighbourhoods that attract people beyond their boundaries will be unfairly portrayed as having higher crime rates.
So, places that attract hundreds of thousands of non-district residents weekly and possess relatively low residential densities – such as Toronto’s financial and entertainment districts – are bound to report much higher rates of assault per 10,000 residents than parts of the city that don’t attract crowds and workers but may in fact be more dangerous.
In recent years both the CBC, the Toronto Star, and a number of other respected media outlets have published maps and stats like these and failed to point out that assaults per resident cannot be used to compare busy downtown wards bustling with outsiders versus sleepy residential enclaves. This is just one of many methodological considerations they have failed to consider for these specific maps.
Urban areas already carry enough crime-related stigma as it is without maps that overstate their crime rates due to methodological faults and limitations. Lists such as the “top 10 neighbourhoods with most crime in each category” based on the CBC map (above) simply cannot be justified given the shaky methodology.
This and other cases of misinformation through the inappropriate assessments of methodological limitations could be avoided by media outlets by having their articles reviewed by competent scientists. Professionals with experience in scientific experimental design would likely spot these issues within minutes and add, at the very least, a few footnotes here and there to save some face.
The most troubling thing about this phenomenon is not that we are failing to adequately educate the public about important issues, but that we are often – inadvertently – spreading misinformation and shaping people’s views with weak or incomplete narratives that may or may not be consistent with reality. We are frequently turning heads in the wrong direction.
By Daniel Arancibia