Lab 4: Housing (Quantitative Data Classification)

Accomplishment: Using data from Statistics Canada, I was able to compare the real estate markets in Metro Vancouver versus Ottawa using various methods of classification, which all drastically altered my results. This exercise proved how easily maps with the same data can be manipulated to drastically cater to different outcomes/visualizations.


Click here to view map.

How different methods of data classification influence the interpretation of data on maps:

In ArcGIS, there are 4 types of classification methods (as defined by ArcGIS):

  1. Natural Breaks: They are based on the best groupings of data with similar values, maximizing differences between classes and data values. They are data-specific and not useful for comparing multiple maps built from different underlying information.
  2. Equal Interval:  This divides the range of attribute values into equal-sized subranges, allowing you to specify the number of intervals, and ArcGIS Pro will automatically determine the class breaks based on the value range. It is therefore best applied to familiar data ranges, such as percentages and temperature.
  3. Standard Deviation: This method shows you how much a feature’s attribute value varies from the mean.
  4. Manual Breaks: This allows you to define your own classes, and you can manually add class breaks and set class ranges that are appropriate for your data.

As a journalist, I would want to choose a classification method that best portrays my case of extreme unaffordability in Vancouver, which I would represent using Manual breaks. Not only does this approach allow me to edit my own intervals, making them easy to follow for the everyday reader, but this also allows me to set my range to argue for a case that the Vancouver housing market is not affordable across the board especially since the class range contains a very small amount of census data falling in the affordable category. Of course ethical implications of following this classification method would involve the fact that I am manipulating data to support my argument, which has specific aims.

If I were a real estate agent however, I would choose the Equal Interval classification method, because due to the scaling, it gives off the impression that real estate in UBC specifically is relatively cheap in comparison to neighboring areas. The range of attribute tables being divided into evened out sub-ranges, therefore the region of UBC that I am promoting to clients appears more affordable, giving them more inclination to purchase homes in the area. Ethically, this appears to be an illusion to clients, as buyers would not actually be able to see how far these UBC properties deviate from the standard, which is purposely altered in my favour as the real estate agent.

 

Housing affordability

The map shown below, which compares affordability in the city of Vancouver versus the city of Ottawa measures household income in relation to owned property costs, using 2011 Canadian Census data. It is calculated by dividing house price by gross annual median household income. Rather than simply displaying housing costs alone, this is a more useful approach to representing ‘liveability’ in the city, as it includes the important factor of incomes which come into play with being able to buy and invest in property. Having multiple variables (housing costs along with median family incomes) puts the Vancouver housing crisis in perspective, and can be useful in explaining phenomena such as outwards suburban migration or homelessness or displacement, and for policymaking decisions.

The Demographia Housing Affordability Rating Survey can be seen as a trusted source for housing affordability rating categories, as it uses a universal “Median Multiple” method to evaluate urban markets. It has been recommended and trusted by the World Bank, the United Nations, the Economist, and Harvard University’s Joint Center for Housing Studies, covering over 400 metropolitan areas globally using the following categories:

  • Severely Unaffordable: Median multiple of 5.1 & Over
  • Seriously Unaffordable: Median multiple of 4.1 to 5.0
  • Moderately Unaffordable : Median multiple of 3.1 to 4.0
  • Affordable: Median multiple of 3.0 & Under

Finally, is this measurement of affordability a good way of determining a city’s ‘livability’?

As stated previously, affordability is one way of determining ‘livability,’ as a way of describing how the costs of living versus personal income and personal expenditures affect ones ability to live in a certain city. However, the term ‘liveability’ is much broader than that, and also takes into account factors like crime rates, quality of education, traffic, air pollution/sustainability, distance to facilities like hospitals or leisure centres, or even personal preferences like living in a quieter suburban area or in a noisy city centre. The list of what makes a city ‘liveable’ is endless, and varies from person to person, making it impossible to determine simply based on level of affordability.

Click to view map.