Detention Logs launched in June 2013, releasing over 7000 incident reports from Australia’s immigration detention centres. These reports are the centre managers’ own recordings of events in detention; from IT failures to riots; from self-harm to births to unauthorised media access.
Since then, journalists have been drawing on the incident reports as an information source in their investigations. They’ve come from a range of publications, covered a range of detention issues, and used the reports in both small ways, to add detail and facts, and large ways, to raise questions and support insight into the immigration detention network.
These investigations have only scratched the surface. They are the starting point for many more stories to come. We want to encourage and assist people using Detention Logs resources in reporting, research, investigation and in ways we can’t yet imagine. We strongly believe that deeper insight into Australia’s immigration detention system is in the public interest.
In line with our project principles, and inspired by ProPublica’s lead, we are providing this guide, a reporting recipe. It walks through methods used by reporters to find insights in Detention Logs data and suggests some new approaches for the future. We aim to expand this guide over time, ideas and contributions are very well.
Approaching detention centre incident reports
First thing, you’ll want to read about the Incident Report data and familiarise yourself by exploring. Once that’s done, one approach to analysing and finding stories in data is by taking near and far views.
The far view
When we take the “far view” approach, we’re looking at the data in aggregate, collections such as; all available reports from Curtin Immigration Detention Centre; every "serious" assault recorded across all centres in 2010 or, every available incident [1.7 MB .csv]. We’re trying to make comparisons between different variable such as time, location or incident type. Are there trends or patterns that emerge?
You’ll want to quickly get your head around the basic analysis tools in Excel, Google Spreadsheets or Open Refine. Don’t be intimidated by spreadsheets, you’ll quickly pick up the basic skills you need.
Sometimes when taking the far view approach we’re just exploring, waiting to see what jumps out at us. More often however, we’re analysing the data in relation to a specific question such as whether a change in policy or circumstance has caused some observable effect. When looking at the incident report data, your questions could include:
- Are different detention facilities characterised by certain incident report types occurring more frequently?
- Has the rate of incident reporting (total number or per capita count) changed over time (across the whole detention network, within geographic regions and at individual centres)?
- Is there a difference in incident reporting between onshore (Australian mainland) and offshore (Chirstmas Island, Manus Island and Nauru [unfortunately the data at Detention Logs does not currently include Manus Island or Nauru facilities]) facilities?
- Do changes in incident reporting occur slowly or suddenly?
- How do the incident reports compare to media reporting of incidents in detention?
- Are responses to policy changes and media reporting observable in the data? i.e. did a change in government policy create a change in incident reporting or a change in the incidents being reported?
Open Refine is an extremely useful tool for breaking down data and looking at trends. Using the raw incident reports dataset [1.7 MB .csv] you can begin to see patterns that emerge by breaking the numbers down in different ways. You can filter by category of incident, timeframe, facility, level and/or do text searches in the “summary” column. The patterns that emerged around reports of violent incidents in particular became key material for a number of reports.
Guardian Australia analysed the data and used it to report on the increase in hunger strikes and self harms in detention centres. The Detention Logs team used a similar methodology to produce a story for New Matilda on the high levels of security failures at Villawood Detention Centre.
When looking for trends and patterns it is often useful to compare the incident dataset with other datasets relating to detention centres or that provide useful comparisons, such as comparable data from other countries or other types of detention systems.
For more tools and approaches for analysing datasets see:
- Database Journalism and Smoking Guns: Finding Evidence, Not Just Anecdote, Steve Doig (Knight Chair in Journalism at the Walter Cronkite School of Journalism and Mass Communication), presentation to The Age newsroom.
- Scraping for Journalism: A Guide for Collecting Data, by Dan Nguyen at ProPublica
- Data Analysis for Policy and Politics, by Edward Tufte
- Drawing Conclusions from Data, by Jonathan Stray
- The Data Journalism Handbook
The near view
Taking the near view is about studying and investigating individual entries in the data: individual incidents and the connections between them.
Is it in the public interest to have more information uncovered about an individual incident? Does the entry expose wrong doing or violations of government policy, the operations of the contract or more broadly the government’s duty of care to detainees?
This approach, investigating an individual or small group of incidents, often starts while you���re actually looking for trends or just browsing through the data; suddenly you come across an incident that seems odd, incomplete or somehow suggestive of a bigger story.
A good way to start is to do a text search of the the raw incident reports spreadsheet [1.7 MB .csv].
When we searched “asbestos”, Excel returned an incident report from 2009 in which a contractor had found asbestos in the mess area at Villawood Detention Centre. Further research into asbestos suggested a follow-up Freedom of Information request be lodged for the hazardous materials register for the centre. After recieving that document, The Detention Logs team published a report with The Global Mail about the high levels of contaminations which led to calls by the NSW Police Association for the safety of the facility to be investigated.
Freedom of Information (FOI) requests are an essential tool for accessing government documents. We used FOI requests to get the full incident report dataset, and many of the Incident Detail Reports that provide more information on individual incidents.
If you would like to get more information out about an incident, you can easily submit a FOI request for its Incident Detail Report by clicking the “Adopt this incident” link near the heading on the incident page.
This link will take you to a page with text explaining the FOI process and a link to a pre-written FOI request for the incident reading “Submit a public FOI request for Detail Incident Report (your incident report number)” The request is submitted through OpenAustralia’s Right To Know FOI system after you go through their brief registration process.
Incident Detail Reports can provide a lot of useful information, such as a list of related incidents, but are often also heavily redacted by immigration officials.
Case studies in detention: How can the data be used as part of a larger investigation or case study?
One of the first incident reports that caught our eye reads, “Client REDACTED attempted actual self harm in SKSA holding rooms”.
SKSA stands for Sydney Kingsford Smith Airport (Sydney International Airport). What did it mean “attempted actual self harm”? What had really happened? What were the procedures around detainee transportation and how had they failed? What duty of care does the immigration department and detention centre staff have to people being detained? The incident summary does not provide answers.
We submitted a FOI request for the Incident Detail Report. The detail report provided more detail, but substancial sections were redacted, raising many more questions about what had happened.
Following the leads in the report we discovered the identity of the man, Ziad Awad. Ziad had been deported to Syrian but was now living in Turkey. Eventually we were able to make contact with him and interviewed him in person in early 2013. The article about his deportation, Deported By DIAC Into Perpetual Limbo published in New Matilda, raised serious questions about the safety and oversight of deportation procedures and exposed inaccuracies in the Department’s record keeping.
The Incident Detail Report also pointed us towards a related incident that revealed details of a controversial restraint technique being used by staff on detainees. We learned more by comparing the policies and procedures at detention facilities with those of the Police and of other government agencies. You can often gain additional insight by analysing incident reports in comparisons with documents from relevant, comparable situations.
Hundreds of the incident reports have the same dry, brutal language that sparked this investigation; “Client REDACTED attempted actual self harm in SKSA holding rooms”. Almost all of them are yet to be investigated—no one person or group can accomplish all this alone.
Follow up on incident reports that grab your attention. It will help uncover insight into how government policy is played out through individual incidents and how it effects individual lives.
Policies and procedures in detention centres
What does the incident report data reveal about how day to day operations at detention centres compare with official procedures and policies? The information in the reports has been used to question statements made by the Immigration Department, revealing inaccuracies, false claims and at times, what seems like an intent to mislead the public.
A strong example is Daniel Flitton’s story in The Age, Officials admit sedative used in detention. It’s been a longstanding policy of both the Immigration Department and Serco that detainees are not to be forcefully medicated or sedated. The incident reports show two clear instances where detainees on Christmas Island had been pinned down and injected with tranquilizers, incidents 1-4FKOKY and 1-6ZY2JS.
Natalie O’Brien produced a report on sexual assaults in mainland detention centres, Assault, gang rape claims in detention rise, for The Sydney Morning Herald following SBS Dateline’s program on assaults at Manus Island immigration facilities. For the investigation O’Brien submitted an FOI request for documents relating to sexual assaults in mainland facilities. In response she received a small number of documents which had already been made publicly available at a 2011 Senate hearing, and nothing else. This was fairly remarkable because a quick scan of the incident reports data at Detention Logs revealed a number of incidents of sexual assault allegations. The Department’s own search of their records in responding to O’Brien’s request, was clearly incomplete.
Beyond the incident reports: how can FOI requests be lodged effectively to the Immigration Department?
One of the key lessons from our investigations was that FOI requests to the immigration department were most effective when we knew what type of document we were looking for. Government department recordkeeping systems vary greatly in their effectiveness, and lodging requests for “all documents” relating to a certain issue will often be ineffective.
One of the ways around this is through documents like contracts that explain the regulatory frameworks for detention centres. New Matilda have published the contracts for mainland detention centres with Serco and the G4S Manus Island contract. Both explain in great detail what other types of documents may exist that can prove useful in investigations.
This is by no means an exhaustive list of the ways the data we’ve published can be used, but we hope it will be a useful starting point. We hope to see the data used as a primary source of how the government records conditions in detention centres, and a secondary source of information about those conditions.
One of the key insights from the data is that the recording of incidents in detention is haphazard, incomplete and prone to error. Always be skeptical of claims made in reports, and consider what may not have been recorded. Understand that these records are made by staff, on the ground, in stressful situations, and that a minority of them are reviewed by managers. Follow up and draw out more information into the public record so that facilities become more transparent to investigation and historical analysis.
Contact Detention Logs if you have any questions or suggestions. You can also get us on Twitter with @detentionlogs.
We hope to expand this guide as the resource grows and is used in new ways over time.