There goes the neighbourhood - Renters in the Census 2016

This week saw the publication of new research from the ANU showing that the problems of housing affordability in Australia don't stem from an undersupply of housing. This is something we've been saying for years - it is not the overall supply that matters, but the kind of supply. Specifically, supply of housing for lower income people.

A few days earlier, the Urban Displacement project in the US updated their San Francisco maps. That project is looking at where people with lower incomes are going when they become priced out of the area they are living in. That updated prompted the crew at #WeLiveHere2017 to ask if anyone was doing similar work here.

@DallasRogers101 @jennacondie is anyone in Australia mapping displacement of low income households? https://t.co/1qDpiIsaSc

— #WeLiveHere2017 (@WeLiveHere2017) November 17, 2017

We thought this was definitely something worth looking into ourselves and we intend to devote a few posts to exploring this issue through crunching some Census data. This first post explores the very lowest income households. The following pictures are here for our mobile readers - they come from an interactive map available here where you can zoom in to any location in NSW and get more details.

We calculated the range of weekly household income for all households in NSW and found the first quintile, or lowest 20% of household incomes, in local areas (for the nerds, 2016 Statistical Area 2) across the state in both 2011 and in 2016.

Then we started looking at how many rental properties in local areas were reporting paying rents that would be affordable to that income level. This includes all forms of public and community housing, as well as private rentals. Using the 30% rule, in 2011 the lowest quintile could afford a property being rented at $155 per week. By 2016 this had risen to $198.30. The following two maps show the raw numbers of properties meeting that number across Sydney. About two thirds of these properties across the state are public or community housing.

Unlike all other rent price sources, such as bonds data and advertisements, the census exclusively measures sitting rents. This is significant in that it explains why some areas may have a higher number of these affordable premises than might be expected if you are used to looking at articles talking about rent movements.

Click the image for a larger picture or the interactive version here

Click the image for a larger picture or the interactive version here

The change is only slight when looking at these raw numbers, but the story becomes much more clear when we look at the change between the Censuses. Across Sydney the proportion of housing available for people on the lowest incomes is dropping - except in a few areas potentially indicating a concentration of this affordable housing.

Click the image for a larger picture or the interactive version here

However, we reckon there's a bit of a difference between an area with very few properties affordable to the lowest income quintile dropping and an area with quite a few affordable properties losing them (or gaining them). To explore that a bit more we've created a final map, which categorises the local areas into 12 groups depending on their placement on a scatter plot. This scatter plot measures on one axis the proportion of housing in the area which on Census night in 2011 was being rented at a rate that was affordable to households in the lowest income quintile and on the other the movement in the proportion of that affordable housing between 2011 and 2016 censuses.

The colour scheme divides those areas losing affordable housing into three equal sets and those areas gaining affordable housing into three equal sets. They are then further divided based on whether they have more or less than the median amount of affordable housing at the 2011 Census.

Once that scatter plot has been mapped we get the following map - zoomed in on Sydney here but the interactive version covers all of New South Wales.
Some of the areas that may seem surprising to appear in the affordable column are there because whilst they are generally affluent areas they do have concentration of public and community housing - or at least did. As one example the Hunters Hill - Woolwich statistical area comes up as affordable due to approximately 167 of the 683 total rental properties in the area being public or community housing - nearly 25%. This is a high proportion considering that across the state only a little over 15% of properties fall into this category.

Click the image for a larger picture or the interactive version here

So what do we learn?

In Sydney it is essentially a bad news story everywhere we turn - either there are unaffordable places getting worse or there are nominally affordable places getting worse. All that bright blue is areas with affordable housing disappearing. The orange is areas with unaffordable rentals that are disappearing. Across the state there appears to be a concentration of affordable housing occurring with most areas falling in the proportion of affordable housing but increases in pockets.

That these rents are sitting rents raises another concern - what happens when people are forced to move? Fortunately a high proportion of these properties are public or community housing but a significant number are in the private market. As such these are households who are in very vulnerable positions. If they do need to move, especially in the private rental sector, they are likely moving on to much higher rents as the market continues to rise.

This is an early version of this data - we haven't adjusted rents and income for household size for instance. It is clear that a single person on the same income as a household of five is more able to fit in a smaller dwelling more comfortably and likely in more affluent areas.

In the next version of these maps we'll be making these adjustments and drawing out the changes in public and community housing as well as looking at slightly higher income groups.