UPDATE: It isn’t really a 299 minute read, but if you choose to explore the map at the end you could well spend 299 minutes.

Australia currently has 53,043 charities registered as at 19 January, 2017. The ACNC is the national regulator of these charities. Of the many things they’re tasked with one of them is to:

maintain, protect and enhance public trust and confidence in the sector through increased accountability and transparency.

One of the ways this is achieved is through the publication of a rich data set each year that shows where charities have sourced their income, and how much has been spent and disbursed to beneficiaries. The ACNC released their most recent data set in mid December 2016. Below are a few observations made by digging into the data set.

The data set

Unfortunately the data set is a little large to display in its entirety, however it is length(acnc_ais) columns wide and contains 45483. This is different to the 53043 number of charities registered as some are exempt from having their reporting information shared. The original data from the ACNC has all sorts of characters in it that complicate things. Unfortunately at this point I’m no wiz with file format conversion even after extensive reading.

The best I could come up with was to attempt to convert the file (I work on OSX) using command line tools (see below) and when I received an error (… cannon convert) go in and isolate the issue in excel and then save again as csv.

iconv -f utf-8 -t iso-8859-1 20170116_Annual_Information_Statement_clean.csv > 20170116_Annual_Information_Statement_Latin1.csv

#> iconv: 20170116_Annual_Information_Statement_clean.csv:44055:79: cannot convert

I worked through this with a combination of deleting entire columns that I didn’t think were too necessary for the initial analysis or editing specific values if they were not a standard ASCII character. Some of the examples I found were an ‘em-dash’, bullet points, and accented letters.

After importing the data I did a few checks on the different sets of columns like staff volumes, income, expenses, etc and noticed that each set of columns doesn’t always tally very well with its relative Total column. Sometimes as many as 2,100 records didn’t balance across a set of data and its total.

The Analysis

After importing the data I did a few checks on the different sets of columns like staff volumes, income, expenses, etc and noticed that each set of columns doesn’t always tally very well with its relative Total column. Sometimes as many as 2,100 records didn’t balance across a set of data and its total.

For the analysis here I am using the calculated totals, not the totals provided as part of the data set.

I thought it would be worthwhile getting a scale of the charity sector. It is worth remembering that this is a reporting data for 45,483 charities not the full 53,043 charities registered with the ACNC.

The total amount of income to these charities in their most recent reporting year is:

Looking at the distribution of income across all the reported charities we see the majority of charities have income around the $100,000 mark.

Further to this though, the ACNC splits the data into different sizes (Small, Medium and Large). It isn’t too evident whether this is done by the ACNC or whether charities can nominate their size themselves. It should be said that even the size data required some cleaning as the values had a mix of cases (e.g. Small & SMALL). The chart below shows the very uneven distribution of charities amongst each size.

It makes sense that you wouldn’t necessarily split the groups evenly and have a third of charities in each group, as this would be an administrative nightmare, however the value designations for each of the bandings could probably do with some realignment. Possibly consider readjusting the Medium group to be charities from $100,000 to $1,000,000.

With a little understanding of the data set and the amount of income realized by these charities it’s worth having a look at the different areas that these charities operate in, and the associated income.

The two largest groups revenue areas are Higher education & Aged Care Activities, however this is across all income streams. It would be worth looking at the same chart, however limiting the income stream to that of Donations and bequests.

While the scale of Donations & Bequests is a fraction of the overall income (less than 10%) it is worth recognising that Australians (Individuals and companies) gave a total of:

This figures under-represents what the ACNC claims as our data is missing about 10,000 charities from the reporting. The ACNC states that donations were about $11.2 billion for the same reporting period.

It’s fair to say that the majority of donations & bequests come from active Fundraising activities, so lets have a look at who the top 10 Fundraisers were and how much they contribute to that $8.8 B.


Looking at the top 10 Fundraising charities two of them don’t have individual giving programs where members of the public can make donations. Thes are The Paul Ramsay Foundation and Rinehart Family Medical Foundation Inc, which appears to be one of several philanthropic vessels for Gina Rinehart.

Further exploration

There is plenty more that can be explored as part of the ACNC data set, however I thought it might be interetesting to leave things with a map below that shows all charities over $1M in donation and bequest income. Enjoy exploring.