For day 3 of dashboard week out task was to tell a story about Australian taxation data. For my story I decided to focus on all the data related to student HELP loans to ultimately find which professions get compensated the most for their degrees.
I wanted as much data as I could get related to HELP loans. The data was scattered across 4 tables and one of the tables related to occupation had all the occupations in codes, so i also needed to convert those ANZCO codes to a more readable format for my dashboard.
- Time series
The time series data had a column for each year, and total number of people with a HELP loan and total amount of HELP debt was in rows, so these needed to be cross tabbed and transposed.
- Age and State data
The Age and state data also included taxable status and gender, however I didn’t need these for my analysis so I had to summarise the data to aggregate all the status’ and gender together.
- Occupation data
The occupations included the most granular classification of occupation that ANZCO has, however I wanted to be able to group all of these occupations into a hierarchical structure. I did this by finding the hierarchy for ANZCO codes else where and parsing out the hierarchy. I also go Help debt by occupation by state from another table in the taxation data.
Now that the data was prepared I was ready to build my visualisation. The first chart shows the percentage of Australians who have a HELP loan in Australia over time. Then we have a break down by age and state, with the ability to swap between average help debt and the percentage of people who have a HELP debt. I also chose to show the unknown and overseas demographics on the map by displaying them as circles next to the map.
One interesting insight that you can get from this section is that out of the 75+ year olds who have a HELP debt that are overseas have an average of $37k in HELP debt owed. Finally we can look at the occupation break down. We can compare the median income with average HELP debt owed for any state or overall. I would have liked to figure out a way to visualise the occupation section with something other than a lot of bar charts, however I still believe this viz gives the ability to find some interesting insights.
You can find the full viz here.