Menu Close
Radio broadcaster Neil Mitchell on Q&A. Q&A

FactCheck Q&A: do refugees cost Australia $100m a year in welfare, with an unemployment rate of 97%?

The Conversation is fact-checking claims made on Q&A, broadcast Mondays on the ABC at 9:35pm. Thank you to everyone who sent us quotes for checking via Twitter using hashtags #FactCheck and #QandA, on Facebook or by email.

Excerpt from Q&A, February 8, 2016.

Now, I’ve never seen so much misinformation as I have on this whole issue. You know, A$100 million a year, I think, the estimate is that it’s costing in welfare payments for refugees. 97% unemployment rate. – Radio broadcaster Neil Mitchell, speaking on Q&A, February 8, 2016.

Amid fierce debate worldwide about how developed nations should respond to people seeking asylum, Australia’s refugee policy has recently had several moments in the global political spotlight – so it’s important we get the facts right.

Broadcaster Neil Mitchell told the Q&A audience recently that refugees were estimated to cost Australia A$100 million a year in welfare payments and that they have a 97% unemployment rate.

Is that accurate? As you’ll see, it’s complicated.

Checking the source

When asked for a source to support his statement, Mitchell sent The Conversation a link to a recent news report in The Australian newspaper, ‘Welfare bill for refugees to top $100m’. The article reported:

Sources say the welfare bill, based on the employment rates of current humanitarian arrivals, is expected to top A$400m over the forward estimates.

Mitchell also sent information from the 2015 Building a New Life in Australia study by the Australian Institute of Family Studies and said:

I confess to one error. I did similar on radio and corrected it. I said 97% unemployed and it should have been 93%. For some reason I had the seven percent employment rate in my head and linked it that way.

You can read Mitchell’s full response here.

Let’s examine the claims one-by-one.

Costing A$100 million a year in welfare?

Mitchell cites a newspaper report as his source for the estimated A$100 million a year figure. That article attributes the number to unnamed sources.

The December 2015 Mid-Year Economic and Fiscal Outlook (MYEFO) includes details on the cost of a plan to resettle 12,000 refugees in Australia on permanent humanitarian visas – in addition to the current humanitarian program intake of 13,750.

You can see here how part of that money is allocated over four years in Appendix A of MYEFO:

Some of the funds set aside for resettlement of Syrian and Iraqi refugees, taken from Appendix A, page 190 of the MYEFO. MYEFO

But it can’t be said for sure how much of the money is set aside for welfare payments. It is likely that many newly arrived refugees will need welfare payments to survive in the short to mid term while they adjust to life in Australia and engage in activities such as English lessons and vocational training, which help them settle effectively. Refugees are subject to the same rules and conditions as Australian welfare recipients and do not receive greater benefits than are available to Australians in need.

It’s also worth noting that Australian policy denied the right to work to some 26,000 people seeking asylum who arrived to Australia by boat after August 13, 2012, when they were released from immigration detention into the community. Many were forced to rely on Centrelink payments to survive.

While the immigration minister began in early 2015 to grant such people the right to work, the process has been administratively difficult and many remain forcibly reliant on Centrelink payments.

So we can’t say for sure if Neil Mitchell is right to say refugees are costing an estimated A$100 million a year in welfare payments. It’s not possible for us to verify the accuracy of a number provided to a newspaper reporter by unnamed sources.

A 97% unemployment rate?

Mitchell has clarified he meant to say 93% for the unemployment rate, and cites the 2015 Building a New Life in Australia study conducted by the Australian Institute of Family Studies (AIFS) as a source for that assertion.

This study aims to gather data over five years on close to 2,400 individuals and families living around Australia, who have been granted a permanent humanitarian visa in the previous three to six months.

The study began in 2013 and data from the first wave of the study are now available.

The study found that of the 2,399 individual study participants

few participants (7% across all respondent types) reported having worked in the last seven days.

So, 100% minus 7% equals 93%. Right? However, it’s not as simple as that.

When contacted by The Conversation, a spokesman for the Department of Social Services, which manages access to the Building a New Life in Australia data on behalf of AIFS, said it was “not possible to accurately calculate an 'unemployment rate’ using the study’s data”.

The Building a New Life in Australia study is not a representative snapshot of all refugees in Australia. As it notes, the majority of people involved in the study were new arrivals: 75% had been living in Australia less than six months and 85% for one year or less.

The study found very high rates of engagement in English language classes and other types of study – meaning that while many of them may not have a job yet, most were working on language and other skills they would need to get one.

Other data on unemployment among refugees

When asked for data showing the unemployment rate among refugees, the Australian Bureau of Statistics referred The Conversation to Table 2 of their Characteristics of Recent Migrants, Australia, Nov 2013 dataset. This publication looks at permanent migrants who arrived in Australia after 2003 and were aged 15 years or over when they arrived.

This data looks at the people in the labour force at that particular time. People not in the labour force are those that are not employed and not actively looking for work. They are not counted in the unemployment rate.

ABS labour stats from 2013 on permanent migrants who arrived in Australia after 2003 and were aged 15 years or over when they arrived. ABS, Characteristics of Recent Migrants, Australia, Nov 2013

To get the unemployment rate from the above, you would need to divide the number of unemployed people by the total number of people in the labour force and multiply by 100 to get the percentage.

For refugees, marked here as “humanitarian”, that works out to be an unemployment rate of about 33%.

Another longitudinal study from 2011 found that while during the early years of settlement unemployment was high among refugees compared to other migrants, 43% of working age refugees remain unemployed 18 months after arrival in Australia.

Also important are the well-documented factors contributing to these higher rates such as lower levels of English proficiency, discrimination, lack of Australian work experience and referees, and difficulty in getting overseas qualifications recognised.

Interestingly, second generation humanitarian entrants have been found to have higher rates of labour market participation than the first generation, and in many cases higher than for second generation Australians.


There were two parts of this statement, one of which can’t be verified, while the other was inaccurate and a misunderstanding of a study’s findings.

Mitchell’s statement that he believed welfare payments for refugees in Australia are estimated to cost A$100 million per year was based on a media report that attributed that figure to unnamed sources. It’s not possible to verify the accuracy of unnamed sources.

The MYEFO does show that more than A$100m is set aside for each year between 2015 and 2019 to help pay for the resettlement of additional Iraqi and Syrian refugees – but it’s not clear how much of that money is for welfare.

As Mitchell himself has said, his reference to a “97% unemployment rate” was an error. He had meant to refer to a 93% unemployment figure, which was his interpretation of the Building a New Life in Australia study showing that only 7% out of the 2,399 mostly new refugees interviewed had worked in the last week.

So is it right to use that study’s data to say there is a 93% unemployment rate among refugees in Australia? No – not according to the Department of Social Services, and not according to the broader range of evidence that we have referred to in this FactCheck. – Lisa Hartley and Caroline Fleay.


I agree with this FactCheck when it relates to unemployment among refugees, but there is some evidence that spending on social security for refugees may be significantly higher than A$100 million.

The statement that refugees have an unemployment rate of 93% is clearly inaccurate, as the the ABS survey Characteristics of Recent Migrants, Australia, Nov 2013 shows.

The unemployment rate is calculated as the number of unemployed as a percentage of the number who have a job plus those unemployed and looking for work.

That means the non-employment rate – i.e those who are unemployed plus those outside the labour force altogether – is 72% for men and 90% for women, giving an overall non-employment rate of close to 81%.

Employment increases over time. So the ABS survey also shows that in 2013 only 19% had a job, but over the 10 years covered by the survey, 36% had a job at some time.

The position on the costs of refugees to the social security system is more complicated.

The ABS survey shows that in 2013 around 59% of refugees and former refugees had government social security benefits as their main source of income.

This was around 54,000 people aged 15 and over, or about 2.2% of the 2.5 million people of working age receiving income support payments in 2013.

Spending on cash payments to people of working age in 2012-13 was close to A$60 billion. If refugees were receiving the “average” payment per person this could actually amount to around A$1.3 billion, considerably more than the A$100 million estimated.

This appears to be something like an upper estimate as it would amount to a payment of around A$24,400 per refugee on welfare. That is considerably higher than most income support payments to people of working age, although refugees who have children will receive family payments on top of their income support payments.

I agree that the level of unemployment among refugees was significantly overestimated by Neil Mitchell, but it is plausible that the total cost of social security benefits for this group is much higher. Definitive estimates are not available.

This does not take into account the taxes paid by refugees with jobs or take account of progression over time, and particularly the contributions that children of refugees will make in future.

For example, a recent study in Sweden found that at a point in time the taxes paid by humanitarian migrants offset about 60% of the benefits they received – Peter Whiteford.

I think this FactCheck is balanced and well written.

Refugees, like everyone else, pay taxes, either directly or indirectly. The longer they stay, the greater their likely contribution.

Unemployment is a very precise technical term. Under the Australian Bureau of Statistics definition, you have to be of working age, out of work and looking for work to be classed as unemployed.

If you are not permitted to work, then you are not unemployed. If you are in education or training, you are not unemployed. If you are caring for young children, you are not unemployed. Neil Mitchell is implicitly comparing apples and pears in his statement.

As the reviewer suggests, a better approach might be to compare the percentage of the working age population in employment. Even then, comparing very recent arrivals with the Australian population would be unfair. It would be better to examine how employment among refugees compares with the Australian population over, say, 10 years, and see how and if they converge. Better still, such an approach might control for factors such as health, disability, age, education and responsibility for children. – Gerry Redmond.

Have you ever seen a “fact” worth checking? The Conversation’s FactCheck asks academic experts to test claims and see how true they are. We then ask a second academic to review an anonymous copy of the article. You can request a check at Please include the statement you would like us to check, the date it was made, and a link if possible.

Want to write?

Write an article and join a growing community of more than 178,900 academics and researchers from 4,895 institutions.

Register now