Written by Isaac Kolamide Ayetoro (4595525), Choon Huei Wen (4595519) & Guligaina Adili (4596420)

The aim of this blog is to present a story about persons who are at risk of homelessness in Australia, specifically persons who are marginally housed in caravan parks in Australia.

HOMELESSNESS IN AUSTRALIA

Homelessness is a continuing public policy issue in Australia and is now recognised as a ‘community problem’. As defined by Australian Bureau of Statistics [ABS] (2012), a person becomes homeless when they do not have appropriate accommodation choices and their current living arrangement: 

• is in an inadequate dwelling; 
• has no tenure, or if their primary tenure is short and unable to extend; or 
• does not allow them to take control of, and establish any social relation of their own.

In order to let people know more about homelessness in Australia, this blog will focus on telling a story about the social characteristics of persons who are at risk of homelessness, specifically persons who are marginally housed in caravan parks in Australia. 


DATA

Data analysis of this blog is done by using SAP Analytics Cloud which able to create a series of visualisations that tell a story about persons who are marginally housed in caravan parks in Australia and at risk of homelessness.

The data used for homelessness analysis is a Census data obtained from Australian Bureau of Statistics [ABS] (2018). All the utilised data for analysis is retrieved from Table_1.2 of the data set, which contains data about selected characteristics of persons in other marginal housing in 2001, 2006, 2011 and 2016.

ABS estimates from Census data for persons who are marginally housed in caravan parks are not classified as homeless, but are considered to be in marginal housing and bearing the risk of being homeless (ABS, 2012). Persons marginally housed in caravan parks are operationalised as those people who were:

• "in caravan, cabin or houseboat in a caravan / residential park or camping ground who reported being at home on Census Night; and where:
• no usual resident reported working full-time;
• the dwelling was being rented for less than $300 per week;
• the landlord was not reported as an employer;
• the dwelling had less than three bedrooms; and
• the combined weekly income of the persons in the dwelling was less than $2,000" (ABS, 2012, p. 41)

When going through the information provided in Table_1.2, it is stated that cells in the table have been arbitrarily adjusted to secure all the confidential data. Consequently, it is also mentioned that cells might not be added to the totals.

It has been noticed that the aggregate of the percentage of persons who are marginally housed in caravan parks per year for most of the data does not sum up to 100 percent. This may largely due to the decimal places of the values or confidential data security.

Hence, to ensure that the data used is consistent and accurate, data cleansing is done by recalculating and standardising the percentage of persons who are marginally housed in caravan parks per year into 100 percent and changed to 2 decimal places instead.

ANALYSIS 1 - AGE GROUPS

Firstly, we started off with analysing the age distribution of people marginally house in caravan park per year by creating stacked bar chart and heat map to analyse the data at a glance.

Visualisation 1: Age distribution of people marginally housed in caravan parks every 5 years

Visualisation 2: Age distribution of people marginally housed in caravan parks every 5 years (in percentage)

Visualisation 1 shows that there were 19,466 people marginally housed in caravan parks in 2001, 12,953 people in 2006, 12,446 people in 2006 and 10,675 people in 2016. In total, there was 45.16% decrease in the number of people marginally housed in caravan parks, from 19,466 persons to 10,675 persons in the 15 year span.

The heat map in Visualisation 2 highlights the two age groups that have the highest percentage of people marginally housed in caravan parks in 2001 and 2016 to show the changes between 2001 and 2016. In 2001, 31.55% of people marginally housed in caravan parks were aged 25-44 but after 15 years in 2016, 42.89% of them were aged 55-74.

ANALYSIS 2 - SELECTED AGE GROUPS

In order to look at the trend of children marginally housed in caravan parks with their parent and the trend of adults continue living in caravan parks, children who aged below 12 and elderly who aged 75 and over are selected to do further analysis. A bar chart is constructed below to show the number of selected age groups of people marginally housed in caravan parks every five years.

Visualisation 3: Selected age groups distribution of people marginally housed in caravan parks every 5 years

From the data in 2001, it shows that there were 2,474 children below the age of 12 amusingly living with their parents or family in housed caravan parks. But 15 years later in 2016, there was a huge 83.63% drop to 405 persons in the number of children below the age of 12 marginally housed in caravan parks and a 38.14% rise from 923 persons to 1,275 persons for people aged 75 and over after 15 years.

This bar chart tells that between the 15 years span, children were more scarcely found or living with their parents in caravan parks or possibly birth control but the adults continued to stay and live there. Also, it illustrates the descending trend of children marginally housed in caravan parks with their parent and the ascending trend of adults continue living in caravan parks.

ANALYSIS 3 - SEX

Moving on, sex analysis is done to analyse whether there are more males or females that marginally housed in caravan parks are at risk of homelessness. A line chart is constructed with a reference line showing the average percentage of males and females who are marginally housed in caravan parks in 2001, 2006, 2011 and 2016.

Visualisation 4: Sex distribution of people marginally housed in caravan parks every 5 years

Gender-wise, an average of 63.94% of male in the 15 years span were marginally housed in caravan parks and the remaining 36.06% were female. Overall, there were more males who are marginally housed in caravan parks and bearing the risk of being homeless.

ANALYSIS 4 - ABORIGINAL AND TORRES STRAIT ISLANDER STATUS

Another characteristic that can be analysed is the Aboriginal and Torres Strait Islander status of people marginally housed in caravan parks. A bar chart is constructed to examine the proportion of Aboriginal and Torres Strait Islander status of persons who are marginally housed in caravan parks. Reference lines are added to the graph to show the average percentage of people with different status in 2001, 2006, 2011 and 2016.

Visualisation 5: Aboriginal and Torres Strait Islander status distribution of people marginally housed in caravan parks every 5 years

Visualisation 5 shows that the percentage of Aboriginal and/or Torres Strait Islander people is stagnant in 2001, 2006, 2011 and 2016 with an average of 5.95%. There were 69.86% of non-Indigenous people marginally housed in caravan parks and at risk of homelessness in 2001, but more than 86% in 2006, 2011 and 2016. At the same time, the percentage of people who did not state their status decreased from 24.64% in 2001 to less than 7% in 2006, 2011 and 2016. This was probably due to people are more open up about revealing their status starting 2006.

Overall, non-Indigenous people were over-represented in 2001, 2006, 2011 and 2016. In the 15 years span data, an average of 83.74% persons marginally housed in caravan parks and at risk of homelessness were non-Indigenous, an average of 10.31% were Aboriginal and/or Torres Strait Islander and an average of 5.95% not stated at all. It can be interpreted that the non-Indigenous people find it difficult to find a proper accommodation which could be due to several factors. It is also an easy way for them not to worry about the usual bill payments and get opportunity to move around freely.


ANALYSIS 5 - STATE OR TERRITORY OF USUAL RESIDENCE

Last but not least, the geographical spread of people at the risk of homeless for people who marginally housed in caravan park is analysed. Visualisation 6 shows the distribution of people who marginally housed in caravan park by state or territory of usual residence.

Visualisation 6: State or territory of usual residence distribution of people marginally housed in caravan parks every 5 years

The red and green bubbles in the geographical map represents the number of people marginally housed in caravan parks in 2001 and 2016 respectively. It can be seen that in every state or territory of usual residence, the number of people marginally housed in caravan parks have a huge drop from 2001 to 2016, this is reflected in the red and green bubble size in the geographic map and the geographical distribution table illustrating the number of people who are marginally housed in caravan parks per year. For all states or territories of usual residence, the green bubbles are smaller than the red bubbles, which indicates that the number of people who are marginally housed in caravan parks and at risk of homelessness is decreased from 2001 to 2016.

Other than that, it can be noticed from the percentage table that in all five different years, Queensland and New South Wales are the top two states that have the highest percentage of people marginally housed in caravan parks and bearing the risk of being homeless. The percentage table shows that in 2001, 2006, 2011 and 2016, more than one-third (35%-37%) of marginal caravan park people were in Queensland, and slightly less than one-third (28%-32%) were in New South Wales.

In summary, two-thirds (64%-67%) of people who are marginally housed in caravan parks and at risk of being homeless are residing in Queensland and New South Wales. 


CONCLUSION

In conclusion, the analysis in this blog indicates that:

1. people who are marginally housed in caravan parks are decreasing since 2006 and the two age groups that have the highest percentage of people marginally housed in caravan parks in 2001 are people aged 25-44, in 2006 and 2011 are people aged 45-54, in 2016 are people aged 55-74;
2. people aged 75 and over and marginally housed in caravan parks are increasing since 2001, while people aged 11 and below and marginally housed in caravan parks are decreasing since 2001;
3. there are more male marginally housed in caravan parks and bearing the risk of being homeless and the proportion of male and female marginally housed in caravan parks are almost the same every five years;
4. nine-tenths of people who are marginally housed in caravan parks and at risk of being homeless are now non-Indigenous people, and the remaining were either Aboriginal and/or Torres Strait Islander or not stated at all.
5. people who are marginally housed in caravan parks and bearing the risk of being homeless are mostly residing at Queensland and New South Wales.

The factors that might account for the decrease in people marginally housed in caravan parks are people have slide into homelessness or survive from financial crisis and found new home (Australian Bureau of Statistics [ABS], 2001).

References

Australian Bureau of Statistics. (2001). Australian Census analytic program: Counting the homeless (No. 2050.0). Retrieved from http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/2050.02001?OpenDocument

Australian Bureau of Statistics. (2012). Census of population and housing: Estimating homelessness, 2006 (No. 2049.0). Retrieved from http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/2049.02006?OpenDocument

Australian Bureau of Statistics. (2018). Census of population and housing: Estimating homelessness, 2016 [Data file]. Retrieved from http://www.abs.gov.au/AUSSTATS/abs@.nsf/DetailsPage/2049.02016?OpenDocument