Archive for ‘Mapping’

October 17, 2009

Toronto's emotional map running hot & cold

Kevin Stolarick, Richard Florida’s “stats guy” at the Martin Prosperity Institute has been up to a bit of mischievous mapping in his spare time.

Using data from a UC Berkeley psychologist who publishes the Big Five Personality Test , Stolarick has mapped out the major emotional of characteristics of Toronto residents by neighbourhood (probably Forward Sortation Areas – the first three digits of a postal code).

The Toronto Star published the maps today: Toronto the Good – and bad and sad and mellow and … .

It’s a relief to see some maps that break the traditional “U” and “O” deprivation patterns. West-enders are extroverted, east-enders are neurotic. Suburban areas tend to be more agreeable, while those along the subway lines are less so. Most of the city is the conscientious type. Those closer to the lake tend to be more open to new experiences.

Now, because the survey is web-based, Stolarick says the sample is probably skewed towards the young (and tech-savvy), but it certainly is a bit of fun!

September 24, 2009

Crime hotspots across Toronto neighbourhoods

(October 29, 2012 Update: CBC release of police crime data by type and neighbourhood)

Today, Stats Can released a hot product: a report on crime in Toronto.  Even though we are one of the safer metropolitan areas on the continent, Neighbourhood Characteristics and the Distribution of Police-reported Crime in the City of Toronto is sure to draw some attention.

Produced by Mathieu Charron at the Canadian Centre for Crime Statistics, the report looks at the location of reported crimes and the characteristics of the neighbourhoods in which they occurred.

The data, drawn from Statistic Canada’s Uniform Crime Reporting Survey (UCR)  “reflect reported crime that has been substantiated by police.” 106,175 incidents were clustered and mapped across the city.

The reports differentiates between violent crime and property crime, finding different correlations. The pattern shows that low-income and nearby neighbourhoods are more likely to suffer spillover effects.

Dividing crimes into violent and property ones, the report found:

  • Neighbourhoods with higher violent crime rates tend to have less access to resources. Education level of residents was one of the best predictors of such access.These neighbourhoods also tended to be “densely populated and have a higher percentage of residents living in multi-unit dwellings” (the tall towers which are the focus of the Mayor’s renewal efforts.) These neighbourhoods are also more likely to have more children, more single-parent families, more renters, and more people of colour.
  • Property crime (theft, break & enter) is concentrated around shopping centres, both large and small, in commercial districts, and in neighbourhoods around these places. Areas with high levels of education or a high portion of manufacturing and office jobs were less likely to report property crime.

Criminologists recognize the spatial patterns of crime. Crime comes in hot spots around the city. Mapping out various criminal activities, the report’s spatial crime patterns follow the same deprivation “U” which marks less privileged areas of the city. Densely populated cores, transportation and shopping hubs, which all draw large numbers of people, tended to report higher crime rates.

The report does not rank or rate specific neighbourhoods, however it did describe “some hot spots…Danforth, downtown east side, and the intersections of Lawrence and Morningside, Jane and Finch, and Jane and Eglinton.”

Here, for those who like the gory details, is what I could see on the maps. The highest levels of crime clustered in the following places:

  • Breaking & Entering: Downsview, Bridle Path, Lawrence Park,Don Mills
  • Drug offense: Jane-Finch, York, Dufferin Grove, Parkdale, New Toronto/Mimico, Trinity-Bellwoods, Regent Park, Greenwood- Woodbine, Crescent Town, Birchcliff, Cliffcrest, Scarborough Village, Kingston-Gallow, Woburn.
  • Major Assault: Jane-Finch, Jane-401, York, Downtown west & east, Lawrence-Kingston Road.
  • Minor Assault: Rexdale, Jane-FinchDownsview, Jane-401, Dufferin-Bloor, Parkdale, Don River-Gerrard, Danforth, Kingston Road, Woburn, Malvern
  • Mischief:  Riverdale, Cabbage Town, York, Morningside/Highland Creek.
  • Motor Vehicle Theft: Etobicoke, Scarborough (where car ownership rates are higher)
  • Robbery: Rexdale, Jane-Finch, Jane-Sheperd, York, Danforth, Woburn
  • Sexual Assault: Rexdale, Jane-Finch, Jane-401, High Park, Bloor-Danforth, Kingston Road
  • Theft: Dispersed along waterfront and main roads
  • Theft from Motor Vehicle: Pearson Airport, Willowdale, High Park, Downtown (west & east), Riverdale, University of Toronto, Scarborough

In contrast, the city’s financial district and the north end of Yonge Street were identified as areas with lower rates of violence. In essence, the central neighbourhoods of the city are higher-income and safer areas, while neighbourhoods with poor physical infrastructure and social resources were more likely to have higher levels of police involvement.

So, the final word probably best belongs to Canadian housing activist Michael Shapcott who wryly noted in his Twitter feed about the study, “Plenty of crime in rich, white neighbourhoods (fraud, tax cheating, ‘white collar’), it just doesn’t get policed/reported.”

read more »

August 8, 2009

Mapping tools add new dimensions to social demographics

Less than a decade ago, easy access to Geographic Information Systems (GIS) caused a paradigm shift  in how we understand demographic data. GIS and spatial analyses have, literally, added new dimensions to our understanding of social landscapes.

Tools to map social data have shifted rapidly through the following stages (note: these are my labels, not some broadly recognized system).

Static maps

Static maps are the ones we remember from our classrooms, hung on the blackboard or tucked into the beginning of our Scholastic Atlases of Canada; inscribed with dozens of symbols which needed to be deciphered with the legends, they covered a range of topics including topographic, climatic zones, agricultural, industry. They were draw by experts.

GIS–enhanced maps

When GIS software appeared, it furnished a way for social scientists to re-examine their stores of demographic data. Instead of comparing along a dimension of time or between similar populations, GIS introduced a way to look at the complex way in which multiple factors overlap and interact within a physical space, the lived world of their “subjects.” GIS capabilities allow social scientists across a wide range of disciplines to add spatial analysis to their analytic toolboxes.

An excellent early example of this stage was The Canadian Council of Social Development and United Way of (Greater) Toronto’s Poverty by Postal Code report in 2004. It looked at the concentration of poverty by neighbourhood, or specifically census tracts, over three decades in Toronto. Professor David Hulchanski’s work through the CURA with St. Christopher’s House on the subject of neighbourhood change and gentrification, has produced similar maps over an even longer time period.

The Toronto Police crime data maps and Toronto Public Health maps do this as well. The maps are static, but the information is conveyed in new and easier to understand ways.

What became apparent from these new analyses is the complex way social problems interact. For instance, Poverty by Postal Code sparked further debate about the importance of neighbourhoods and place-based strategies. United Way and the City established the Strong Neighbourhoods Taskforce, which by mapping proximity to service against social need, sparked new planning priorities.

Web 1.0 maps

Web 1.0 maps moved mapping off computer desktops and onto the internet, allowing broader interactivity. With Web 1.0 technology, viewers are able to move through pre-mapped⁄pre-coded data to find answers (sometimes) to their own questions. Good local examples of these are:

  •’s Close to Home maps of 211 Ontario data, allowing newcomers to search for services closest to their residence/place of work.
  • City of Toronto developed MapIt, an interactive map which allows viewers to select what city services should be shown on the map and then to zoom to an area of interest.

Statistics Canada data has been incorporated into several Web 1.0 vehicles to make accessing it more interesting than looking at a set of dry tables. Several Canadian examples exist, and many of these are incorporating other data sources as well:

  • The Canadian Council on Social Development has established a national platform through its data liberation initiative for municipalities and non-profit agencies. The Canadian Social Data Strategy has a public front door and an area for local agencies to have access to further data.
  • Although requiring registration and log-in, the Canadian Mothercraft Society has also built a very usable platform for community agencies to select and map out data in their areas of interest.
  • The Government of Newfoundland & Labrador led Canadian provinces in establishing Community Accounts, a web-based map system which produces local profiles upon a range of factors which may be selected by the site visitor. Nova Scotia has followed suit.
  • The Toronto Star has a blog and staff dedicated to mapping newsworthy social issues.
  • Using a democratizing Google mash-up, the creative Baby Name Map was established in Calgary.

Web 2.0 maps

Web 2.0 mapping is taking GIS interactive. (Web 2.0 engages internet surfers in two-way information exchanges, so that they can add information as well as get it.)

I have been able to identify several ways this is done in mapping:

Open Source GIS: The power of mapping technologies has, in this initial period, remained concentrated in the hands of experts who have access to software which can cost thousands of dollars. Several open source software are emerging and refining to the point that GIS software will become more available to everyone. Grass is one of the most preeminent ones. My Maps on Google Maps also give easy access to people to map their own worlds.

Crowd-sourcing: This method farms out work, realizing on the small contributions of many to make sense of complex problems. For instance, Industry Canadainvited Canadians to submit information about their broadband access which could then be mapped out across Canada to identify areas with significant service gaps.

Community mapping: Google maps are some of the frequent examples of interactive mapping. Family Service Toronto is working with Waterloo’s Comap to launch a community mapping initiative in the Teasdale-O’Connor neighbourhood, which will invite local agencies and residents to contribute and shape the maps of the neighbourhood.


Real-time: Real-time mapping is still emergent. For example, an iPhone app uses GPS to update your location to selected friends and family.  Twittervision and celebrity-stalking websites like Gawker’s Stalker are powerful because they add a geographic scale to the information shared.

Other good examples and methods are continuing to emerge. Please feel free to share other good examples!

read more »

July 18, 2009

Mapping jail and university admissions

The results are in from the stellar Toronto Star team again. This week-end, they released two sets of maps, in many ways the obverse of each other:

The latter map is the result of a court order, as described in a previous post and a strong contribution to  the argument for place-based interventions. Our thanks to them.

The maps looking at university admissions also support the work being done by the Toronto District School Board’s researchers who have mapped university applications and other academic indicators by neighbourhood.

These unsettling maps lay how applicants to one of the most prestigious universities in Canada live in different worlds than the the places where people are being jailed. Opportunities are literally mapped out.

The co-incidental (?) and simultaneous release of maps is evocative of the statistic that, in many American inner cities, there are more young men in jail than in college or university.

I’ll look at more of the details in these maps in another post.

June 11, 2009

A reminder: CIW requires rigor in its statistics

My favourite statistical test emerged out of one of consultations for the Canadian Index for Wellbeing (CIW):

The Chief Statistician of Newfoundland and Labrador, Alton Hollet, explained some of the rigorous testing used in the development of the province’s Community Accounts, which allows people to search for data about their local community. The user-friendly website is the best demographic profile tool in Canada.

After telling the story of how a few audience members had fallen asleep during the initial and very dry presentations of the Accounts, Hollett sternly warned the gathered experts that people must see their lives reflected in the statistics to be presented to them because “if you can’t see yourself in the mirror, how are you going to comb your hair?”

It’s a statistical test — and a laugh — that has stayed with me.

June 1, 2009

2006 Toronto demographics

Map of Toronto

Image via Wikipedia

If you are feeling like a real demographic data geek, you’ll enjoy this presentation titled, Demographics of Toronto: A visual presentation of population sub-groups. It was assembled for a Toronto consortium meeting of the Canadian Social Data Strategy by Harvey Low, the City’s planning analyst who convenes the group of community agencies and city divisions.

Topics include 2006 census variables on population and age variables, immigration, and income. There is also a draft of a “mosaic indicator” to measure diversity and the degree to which Toronto neighbourhoods are ethnic enclaves.

If you are looking for more specific data, take a look at the links on the side of this blog. They include links to a range of sites. Particularly interesting is the City of Toronto’s Toronto-wide and neighbourhood level data across a range of domains.

read more »

May 28, 2009

The strength of EDI as a predictive tool

I am probably about to commit heresy: I hate the EDI.

The EDI, in long form the Early Development Instrument, has gained popularity as a population tool to rank students’ readiness for school. Developed by Dan Offord and Magdelena Janus at McMaster University, and popularized by Clyde Hertzman at the University of British Columbia, the EDI has been shown to have a strong correlation to the likelihood of a student cohort to achieve academically. But more tellingly, it strongly correlates to students’ socioeconomic backgrounds and the neighbourhoods where they live.

The EDI tool is administered in Senior Kindergarten by classroom teachers in the space of about fifteen minutes per student. To quote, students are assessed on five domains:

  • Physical Health and Well-being referring to physical readiness for the school day, physical independence, and gross and fine motor skills.
  • Social Knowledge and Competence referring to overall social competence, responsibility and respect, approaches to learning and readiness to explore new things.
  • Emotional Health and Maturity referring to prosocial and helping behaviours, anxious and fearful behaviour, aggressive behaviour and hyperactivity and inattention.
  • Language and Cognitive Development referring to basic and advanced literacy skills, interest in literacy/numeracy and memory, and basic numeracy skills.
  • Communication Skills & General Knowledge referring to the child’s ability to communicate needs and idea effectively and interest in the surrounding world.

In each of these domains, children who come from tougher economic circumstances or from outside the dominant linguistic or ethnoracial group are invariably disadvantaged. For instance, whether a child arrives to school with appropriate clothing, can discuss an idea, or knows her way around a picture book are all EDI measures. Low EDI scores are often, though not always, evidence of deprivation.

Children who start from further behind also face a higher hurdle, if they are to measure up to their peers. A different framework would  measure the improvements children may have made; instead the EDI uses a threshold to measure a child’s readiness for school. You make the cut-off or you’re at-risk.

The strengths that marginalized students might bring to the classroom, but which fall outside the scope of the the framework for “school readiness,” are also not recognized in the same weighty way. Internal resiliency in the face of a strange school setting doesn’t get measured.

(My favourite example of this was when the girl from across the street who spoke English as a second language began kindergarten with my daughter, she relied on my daughter to repeat the teacher’s instructions slowly. Then, in the afternoon, when they both attended Heritage language classes, they reversed roles. It was a creative coping strategy.)

Finally, I hate the the EDI because it can act as a proxy for teachers’ middle class prejudices and ethnocultural biases. However, therein also lies its strength.

Not surprisingly, groups of students who do poorly in academic rankings in Kindergarten generally continue to do poorly in the eyes of their teachers in higher grades. Teachers are at least consistent.

Like Robert Fulgham’s Kindergarten Poem, all your school really needs to know is how you did in kindergarten. If your Kindergarten teacher thought you were unruly and inattentive, then probably so will subsequent ones.

Research has shown that the EDI is a reliable predictor of children’s likelihood of completing school successfully. We can tell that early on who might not make it.

The EDI’s predictive ability is sad confirmation of the social gap that some identifiable demographic groups come from further behind and stay behind throughout their schooling.

Yet, there is hope. Hertzman’s work, with the school system in British Columbia, shows that a coordinated, community-based response can make a difference in the school readiness of all children. So, that is where our work begins.

March 5, 2009

Are hospital visitors targeted for parking violations, or are we just negligent roadhogs?

My parents go to the hospital so frequently that the last time I escorted them, they carefully coached me in how to avoid getting a parking ticket. It’s energy well-spent, given the frequency with which parking tickets are handed out around hospitals from York region, to Ottawa, from Newfoundland to Australia.

Whether you are visiting, attending a doctor’s appointment, or rushing there for an emergency, parking tickets are a common part of the hospital experience, along with high parking fees, shortages of spots, and meters which expire in short intervals.

A recent piece in the Toronto Star highlighted how frequently hospitals visitors are stung by the green hornets here in Toronto. The streets around hospital made up half of the top ten sites for parking tickets in 2007. The Ottawa Citizen found similar patterns in their examination of the issue in 2007. The Vancouver Sun also found the same, to a lesser degree.

It’s the sort of thing that drives people crazy, filling Bulletin Boards and other blogs (see here for a hilarious list of the ten worst parking tickets ever issued).

Some places in the world are trying to find a solution. Scotland now offers free parking at most of its hospitals, and Wales is considering the same, while recognizing the complexity of such an endeavor, and wondering how to discourage “freeloaders” without setting up another expensive bureaucratic layer.

Some argue, perhaps fairly, that if you own a car, you need to take responsibility for it. Residents who live near hospitals have to put up with slackers on a daily basis. It must grow tiresome.

However, hospitals are one of the likeliest places in the city where some administrative discretion should be used. People attending hospitals are often ill, or escorting those who are, and they have little control over the sorts of delays they may face once inside.

I have my bias in answer to the question: I remember a sweet and random act of parking kindness  I received at my local hospital once, when I raced, daughter in my arms, son at my side, into the emergency room. When we left, all safe a few hours later, I realized that I had parked by the entrance and not even noticed the meter by my car. But there sat my car ticket-less.

Someone had put some money into the meter.

read more »

February 13, 2009

Ethnic enclaves in Toronto, 2001 – 2006

A packed house gathered last week at the Joint Centre for Excellence on Research in Immigration and Settlement. Building on their earlier work on ethnic enclaves in Toronto, professors Mohammed Qadeer (from Queen’s) and Sandeep Kumar Agrawal (from Ryerson) were speaking about the residential patterns of seven ethnic populations in Toronto:  African Blacks, Caribbean, Chinese, Italian, Jewish, Portuguese, and South Asian. (These are as reported at the census tract level by individual respondents to the 2001 and 2006 censuses living in the census metropolitan area (CMA) of Toronto.)

While the maps in their PDF presentation were the most interesting, Qadeer and Agrawal also laid out a few key elements about “ethnic enclaves”:

  • Enclaves are defined as residential concentrations with supporting cultural institutions and services.
  • Enclaves are distinct from ghettos because they happen through a positive choice, rather than a lack of choice. Measuring this, however, is a challenge.
  • Enclaves are an important step for Canadian newcomers on the way to settlement and integration.

Using GIS analysis, Qadeer and Agrawal’s found that ethnic enclaves are extending (so that they are now more widespread) and consolidating (single ethnic groups were more likely to be a higher portion of a neighbourhood). This growth, they found, was often spurred by new immigration.

However, they also found a wide variation in the likelihood of people of various ethnic groups to live within their own neighbourhoods, and that no enclaves were exclusive. All city census tracts had some ethnic mix.

The study provokes further questions to explore, many which were asked that afternoon. Further research could be done to look at these trends over a wider range of years and among other Canadian geographies or at alternate geographic levels (dissemination areas instead of census tract). Also worthwhile would be a examination of the shifting residential patterns of the City’s largest ethnic group, those of British ancestry, and, and more compelling, whether there is a tipping point when “white flight” becomes a reality.

Finally, the study gave a quick look at the percentage growths among various ethnic groups; Russian and Ukrainian populations grew the most quickly. The largest groups are British, South Asian, and, then, Chinese. (For more a detailed description, see the City of Toronto’s 2006 census broad overview and the profiles of specific ethnic groups.)

read more »

October 21, 2008

Racial divisions tracking income polarization

Three recent learnings from the Ontario Nonprofit Housing Association conference scared me about the future of our city:

I was sitting through a presentation I had seen a few times, about the growing concentrations of poverty across the city and the high income enclaves that were also emerging, when I was struck to hear how substantially these aligned with the emerging racial division in the city. Just as income has polarized even over the past five years, so have the racial divisions. Neighbourhoods which were mainly white at the 2001 census are now likely to be even more white, according to research being led by Professor David Hulchanski at the University of Toronto.

So, the next morning, as I sat through an anti-racism workshop at the same conference, we were asked if we saw evidence of racism in our communities. Hulchanski’s work shows that, as the city’s foreign-born population now hits 50% of residents and people of colour will soon be a majority of the population, many white people, especially those living in high income areas, are less and less likely to have contact in their day-to-day lives with those from another racial background.

Finally, in another session, we talked about the dynamics of what happens when mixed neighbourhoods disappear. Like the idea of supermajorites, as described by political scientists, when populations become more homogenous and ideas and social mores are not challenged, they tend to become more extreme in their positions.

All this means that urban residents, living within increasingly racially and economically segregated neighbourhoods, will become increasingly isolated and separate in their world views and experience.

As I said, scary.

David Hulchanski’s work can be found at maps of city neighbourhoods with very high concentrations of white and visible minority populations and a recent presentation.

<span>%d</span> bloggers like this: