While corrections departments around the country report that more than 80 per cent of people in prison are working, in reality only about half actually have employment while in custody. This is because of how these departments are reporting the rate of employment and who is counted as being “eligible”.
I waited in line at reception after entering custody in 2015, and listened to the guard asking the standard 20-odd questions for new arrivals with the men in front of me.
“What’s your highest level of education?”, he asked in a tone that clearly indicated he was as bored as I was.
“Bachelor’s degree”, I replied.
He stopped typing, actually looked at me, and eyed me up and down for a few seconds, probably trying to figure out if I was screwing with him or not. Eventually, he must have decided I was telling the truth. He picked up his phone and dialled an extension.
“Hey Anthony, you gotta come down to reception and meet someone here,” he said.
As I soon found out, Anthony was the education officer, and he quickly asked me to clarify my education level.
“This is great,” he said. “We’ve been looking for a new education sweeper for a while. Most of the guys here aren’t very literate so they can’t do the job. How would you like to be the sweeper?”
“What’s a sweeper?” I replied.
Among the many things I learnt on the first day of my imprisonment was that ‘sweeper’ was prison slang for an employee. I accepted the role. In exchange for working in the education department, my prison ‘pay’ was increased from the base rate of about $15 a week to $23, about an extra 40 cents an hour.
Despite the ridiculously low pay, it wasn’t until I had been in for a few days that the other prisoners explained how lucky it was for me to have a job. You see, there were about 30 prisoners in the main wing, and alongside the person who washed the floors, I was only the second one to be employed.
So imagine my surprise when I read the Corrective Services NSW 2024-25 annual report as part of my PhD research more than a decade later.
The ‘Work and Education’ page noted their ‘key highlights’, which included that 85.7 per cent of ‘eligible inmates in custody’ were working.
I did a double take. More than 8 in 10 prisoners working? I thought there was no way that was correct.
Then I noticed the keyword: ‘eligible’. But the annual report didn’t clarify who was eligible and who wasn’t.
Thankfully, the Productivity Commission (PC) does.
The PC is an independent government research body. Every year it releases a Report on Government Services, including a section on Corrective Services.
This includes data on many things including operating costs, prison population, education rates in custody, time out of cells and employment. The PC said there are 11 criteria that make a prisoner ineligible for work. Some of these, such as “ill health”, “hospital patients” and “aged prisoners who are unable to work”, are entirely understandable. Others less so.
Prisoners who are deemed ineligible include remandees who choose not to work and prisoners who are “at centres where the jurisdiction's policy is not to provide work or where work is not available.”
Yes, you read that correctly. You can be deemed ineligible for work in prison, even if you want to work, if the prison does not provide work opportunities for you.
When I was in custody at the reception prison, there were only two jobs available for the 30 prisoners in the main wing. This is an employment rate of 6.7 per cent. Many of these prisoners were on remand but they were not actually asked if they wanted to work at the prison or not because there was no work available.
Because there was no work available, it appears corrections would have counted 6.7 per cent of prisoners actually working as a 100 per cent employment rate.
The PC clarified that the national rules for counting data have been agreed upon by state and territory correctional agencies, rather than determined by the PC itself.
The PC was unable to tell me how many prisoners were actually working as correctional agencies chose not to give them this information. Correctional agencies only provide data on what percentage of prisoners they have deemed eligible to work, using their own criteria. This data gives no indication of how many prisoners were deemed eligible.
Realising this incomplete data is not helpful in understanding actual employment rates in custody, I began seeking out the missing information myself.
Corrections in most jurisdictions freely released the relevant data to me, though South Australia, Queensland and the Northern Territory refused and so I went through Freedom of Information.
The Northern Territory application has yet to receive a response. The following table shows the percentage of eligible prisoners working – the data that corrections provides to the PC – next to the percentage of prisoners who are actually working.
Among the Corrections departments that have provided the data, it shows that on average the actual number of prisoners working is more than 25 per cent lower than the figure reported to the PC.
Associate Professor Benjamin Spivak is an expert in statistics at Swinburne University, and he shares my concerns about how Corrections choose to report employment data.
“It means that if a person is held somewhere where work is not available they are not counted in the rate,” Spivak said.
“This would be the equivalent of reporting unemployment rates for Australia and not including people who could not find work.
“The rate that is reported is useful in telling us how many prisoners decline work that is available, but doesn’t tell us how many prisoners wanted work but could not get it. An easy fix is for reporting to include how many people are in custody, how many are working and how many were classed as ineligible and why.”
Dr Matt Maycock, a senior lecturer in Criminology at Monash University who has written about prison labour around the world, also has concerns.
“The official figures in relation to work in prison are problematic,” Maycock said.
“Given that remand prisoners are not compelled to work, and are not included in these figures despite constituting 42 per cent of prison population.”
Spivak also notes an issue with Corrections not counting remandees who do not work in their data, while counting remandees who do work. It also remains unclear how many remandees actively chose not to work, or how many were simply classified as choosing not to work by default because there was no work for them to choose from in the first place.
After reporting my concerns, the PC has told me that it will take my feedback on board for how data is collected for the 2028 Report on Government Services, as it is too late in the data collection process to change how data is reported for 2027.
Corrections in most jurisdictions freely released the relevant data to me, though South Australia, Queensland and the Northern Territory refused and so I went through Freedom of Information.
The Northern Territory application has yet to receive a response. The following table shows the percentage of eligible prisoners working – the data that corrections provides to the PC – next to the percentage of prisoners who are actually working.
Among the Corrections departments that have provided the data, it shows that on average the actual number of prisoners working is more than 25 per cent lower than the figure reported to the PC.
Associate Professor Benjamin Spivak is an expert in statistics at Swinburne University, and he shares my concerns about how Corrections choose to report employment data.
“It means that if a person is held somewhere where work is not available they are not counted in the rate,” Spivak said.
“This would be the equivalent of reporting unemployment rates for Australia and not including people who could not find work.
“The rate that is reported is useful in telling us how many prisoners decline work that is available, but doesn’t tell us how many prisoners wanted work but could not get it. An easy fix is for reporting to include how many people are in custody, how many are working and how many were classed as ineligible and why.”
Dr Matt Maycock, a senior lecturer in Criminology at Monash University who has written about prison labour around the world, also has concerns.
“The official figures in relation to work in prison are problematic,” Maycock said.
“Given that remand prisoners are not compelled to work, and are not included in these figures despite constituting 42 per cent of prison population.”
Spivak also notes an issue with Corrections not counting remandees who do not work in their data, while counting remandees who do work. It also remains unclear how many remandees actively chose not to work, or how many were simply classified as choosing not to work by default because there was no work for them to choose from in the first place.
After reporting my concerns, the PC has told me that it will take my feedback on board for how data is collected for the 2028 Report on Government Services, as it is too late in the data collection process to change how data is reported for 2027.
