The Immigration Question We've Stopped Asking

On 12 May 2024, in a Brisbane suburb called Acacia Ridge, a man named Emmanuel Saki stabbed another man, Bosco Minyurano, to death.

Saki was in Australia because an immigration tribunal had restored the visa the government had cancelled. He had a serious criminal record. Section 501 of the Migration Act lets the government cancel a visa for that, and Saki's had been cancelled. The tribunal agreed the cancellation was lawful. It set it aside anyway, because a Ministerial direction called Direction 99 told the tribunal to give primary weight to a person's "ties to Australia during formative years". Saki had those ties. The visa came back.

He was one of more than thirty people whose cancellations had been reversed in the same way. The unit meant to monitor them had not been staffed. The Minister had not been briefed. Weeks after the killing, the Secretary of Home Affairs told a Senate committee that the department had "failed". Direction 99 was rescinded the next month and replaced with Direction 110.

This essay is not about Emmanuel Saki. It is about the machinery that failed around him, and it is about a larger question the Saki case is only one symptom of. Australia's migration system has not been audited publicly for fourteen years. It is overworked, under-resourced, and asked to do work it was never designed to do. It enforces at the back end what it no longer properly assesses at the front. And on the biggest question of all – who we let in, and on what basis – it has quietly stopped asking.

I. The numbers, and why this essay parks them

The Australian debate on immigration is obsessed, not unreasonably, with the number. Net overseas migration was 446,000 in 2023-24, after peaking at 536,000 the year before. In the two years to mid-2024, Australia's population grew by about 1.54 million people – roughly the population of Adelaide – and four-fifths of that growth came from overseas migration. One new home was built for every 2.1 migrants in the year to September 2024. Rents rose faster than in a generation.

Whether that inflow has been too fast, too slow, or about right is a serious debate. It has real consequences in hospitals, schools, roads, the rental market, and the wages of the people Australia already has. It is a debate this essay is deliberately not going to have. It deserves its own essay, and will get one.

The point of this essay is that there is a second debate, of equal weight, which has gone almost entirely unattended while every political argument was being fought on how many. It is about the quality of the question. The people we let in – the families we reunite, the workers we sponsor, the students we accept, the relatives they bring with them – we are asking shallower questions than we used to, with less care, less follow-up, and less resource. The machinery that does the asking has broken in quiet, specific, documented ways.

Unless that machinery is fixed, no decision on numbers – up, down, or held – will produce the result the country is hoping for. A question seriously asked is the foundation. The arithmetic sits on top.

II. Busy at the back, shallow at the front

The character test, section 501 of the Migration Act, is the part of the system the public hears about. It is the instrument that lets the Minister cancel a visa on character grounds, usually after the holder has been convicted of a serious offence. Cancellations under section 501 rose from about 76 in 2013-14 to 948 in 2024-25, most of them mandatory under the scheme introduced in 2014. This is real work, done hard.

The work upstream is where the system has thinned out.

The Australian National Audit Office has not publicly audited the immigration security-assessment function since 2012. The 2012 report, Audit Report No. 49, documented staffing shortfalls of up to 31 per cent and backlogs of up to 918 days. The Commonwealth Ombudsman reported similar problems in 2016. The Administrative Review Tribunal is currently about 100 full-time positions short of its approved budget, with 46,600 student-visa appeals in its queue. ASIO's visa-assessment workload is now being done alongside a new positive-vetting function handed to the agency in 2023, at the same time its counter-terrorism load has grown.

Abul Rizvi, a former deputy secretary of the immigration department, is the most credible insider voice in the debate. His diagnosis is blunt. Enforcement has been misdirected. Section 501 is being used prolifically against prisoners, where the cases are easy and the political optics flattering. Meanwhile, labour-trafficking and sham-sponsorship scams have industrialised in the blind spots, where the law is complicated and the vetting is thin. The machinery is being run hard in one direction, and left to rust in the other. Rizvi is right.

III. What Singapore asks, what Denmark does

Singapore is the country to watch.

In September 2023, Singapore introduced the Complementarity Assessment Framework – COMPASS. Where Australia's Core Skills Income Threshold is a single wage floor ($73,150 from July 2024), COMPASS asks six questions. Three are about the applicant: salary, qualifications, and whether the applicant's skills match a national shortage. Three are about the sponsoring firm: the diversity of its professional workforce, its support for local employment, and its contribution to Singapore's strategic economic priorities. An applicant must score forty points across the six to qualify.

The most ambitious of the six is C3: the nationality concentration of the firm's professional workforce. A firm whose senior professionals are dominated by a single foreign nationality scores lower on C3, and must score higher elsewhere to clear the threshold. C3 does not cap any nationality. It does not tell a firm whom to hire. It asks the firm to demonstrate that its sponsorship is adding plurality to its workforce, rather than concentrating it. Early data from Singapore's Ministry of Manpower shows the share of firms with single-nationality-dominant professional workforces has fallen seven per cent since the rule came into effect.

This is the design Australia should copy. C3 catches the actual pathology in our employer-sponsored streams, which is not low pay. It is sponsors who build ethnically captured workforces around themselves: opaque to inspection, and uniquely able to exploit their own people.

Denmark is the second example worth studying, for a different reason. Under the Social Democrat government of Mette Frederiksen, Denmark has pursued what the party calls, plainly, "zero asylum". The policy includes a cap on non-Western immigration, a requirement that immigrants work at least thirty-seven hours a week before becoming eligible for welfare benefits, and the revocation of residency permits for some Syrian refugees on the grounds that parts of Syria are now safe to return to. Frederiksen's case, made before she became prime minister, was unapologetic: the price of unregulated mass immigration, she wrote, is paid for by the lower classes.

What is striking about Denmark is not the particular policies. One does not have to agree with every element. What is striking is the intellectual honesty. Denmark measures, publicly, which inflows have contributed to the country's economy and culture and which have not. It categorises by source country, by education level, by labour-market outcome. It then designs policy on the evidence. A country that refuses to categorise has refused to evaluate. It is asking its migration program to be immune to honest audit. No serious institution can be.

IV. Sweden, and the lesson Britain has just learned

Sweden is the cautionary tale.

Over twenty years of high immigration, weighted heavily towards refugee flows from non-Western countries, Sweden has recorded a measurable and persistent difference in crime rates between immigrant and native-born populations. The Swedish National Council for Crime Prevention has found that people of foreign background are suspected of crimes at roughly two and a half times the rate of the native Swedish population. Over time, first-generation immigrants converge somewhat with the native rate. Second-generation immigrants do not. In recent data, the gap between the second generation and the native rate has widened, not narrowed.

Socioeconomic factors – unemployment, poverty, language, education – explain most but not all of the difference. The point is not that immigration causes crime. The point is that a country that does not integrate well, and that does not vet for integration capacity in the first place, finds itself thirty years later explaining gang violence in its quiet suburbs. Stockholm, until recently one of the most orderly capitals in Europe, now records more fatal shootings per capita than many American cities. Denmark noticed. Sweden's neighbours have followed Copenhagen's policies, not Stockholm's.

Britain's warning is different, and recent. Between 2020 and 2024, the United Kingdom opened an uncapped Health and Care Worker visa route. Main-applicant grants hit 146,477 in 2023, up 91 per cent on the previous year. By early 2025, Britain had revoked more than 470 sponsor licences in the care sector, stranding 39,000 workers attached to sponsors who should never have been certified. The student-dependant route ran the same pattern: applications rose from about 16,000 in 2019 to 143,595 in 2023, as a route designed for the families of genuine students was used as a work-rights backdoor. The UK restricted dependants in 2024 and applications fell to 22,000 immediately.

Britain did not do anything malicious. It opened a route, assumed the sponsor-side checks would hold, and discovered, retrospectively, that they had not. Australia has watched this happen twenty time zones away. It has not yet absorbed the lesson.

V. The question Australia has stopped asking

Behind the mechanics of the machinery sits a harder question. What kind of values compatibility does a country of laws ask its newcomers to demonstrate?

Australia's citizenship pledge includes a promise to uphold Australian law as supreme over any other source of legal obligation. That is a values test, however lightly worn. It is also a test easier to sign than to believe. A British ICM poll in 2016 found that forty-one per cent of British Muslims supported the introduction of particular aspects of Sharia law alongside or in place of British law. A 2006 NOP survey found that twenty-eight per cent wanted Britain to be an Islamic state. No equivalent survey has been undertaken in Australia, and the absence is itself a political fact.

This is a question a serious country has to be willing to ask. Not in broad indictment of Islam, which would be wrong and counter-productive. In specific terms: Does the applicant accept the supremacy of Australian law over religious law? Does the applicant recognise the equal standing of men and women, gay and straight, Christian and Muslim and Jewish and atheist? Does the applicant reject political violence? These are values questions, not ethnic ones. A serious migration program asks them of everyone.

The reason Australia has stopped asking is not that the questions became unaskable. It is that the political cost of asking them – the reflex accusation of racism, the labelling of any values inquiry as bigotry – became too high. That cost must not be allowed to set policy. National interest has to be looked at with clear eyes. Performative virtue, at the expense of a functioning migration system, is not a form of virtue. It is a form of negligence.

There is a second version of the same problem. Elected officials at state and federal level have, at various times, campaigned inside specific diaspora communities with material promises: faster family-reunion processing, cultural or religious carve-outs from general rules, visa-class expansion targeted at particular source countries. Much of this is not illegal. It is also not honest. A migration system that is quietly arbitraged for ethnic-bloc votes is not a migration system. It is a marketplace, and like all marketplaces it rewards the organised at the expense of the country the organisation was built inside.

Values questions are hard to ask well. Australia has asked them badly before – the Dictation Test of the White Australia era was a racial filter dressed as a language test. The lesson is not that values questions are dangerous. The lesson is that they must be asked properly, on clear grounds, with rule-of-law discipline, and of everyone. A serious country does not stop asking. It asks better.

VI. The argument against

The case for leaving the system roughly as it is, or lightening it, is not made by activists. It is made by serious researchers. It is also, however, made by people who are not politically neutral. Grattan is a policy think tank that leans progressive-technocratic. The Kaldor Centre at the University of New South Wales is an institute dedicated to refugee law. The Law Council of Australia is, by institutional disposition, on the side of procedural fairness for the applicant. Their case deserves a fair hearing. It also deserves to be read with awareness of where they stand.

It comes in three parts.

The fiscal case. Grattan estimates the lifetime net fiscal contribution of a skilled migrant at around $55,000. That number sounds like an argument until it is examined. The average Australian worker pays roughly $20,000 in income tax every year. A lifetime net contribution of $55,000 is less than three years of the average worker's tax, spread across a forty-year working life. It is thin.

More importantly, it is an average. Denmark has done the disaggregation Australia has refused to do. In 2018, the Danish Finance Ministry reported that immigrants from the Middle East, North Africa, Pakistan and Turkey cost the state about 85,000 kroner per person per year. Immigrants from other non-Western countries cost about 4,000 kroner per person per year. Immigrants from Western countries, by contrast, contributed roughly 47,500 kroner per person per year in the other direction – about 14 billion kroner in aggregate across the Western-origin population. The variation was enormous. The Grattan headline is not a finding. It is an average that hides the question the evidence could answer, if Australia chose to ask it.

The demographic case. Australia's total fertility rate fell to 1.48 in 2024, the lowest since 1929. On the standard argument, high migration is the fix. Dick Smith has spent a decade making the simpler counter. Migrants grow old. A migration-driven ageing fix is a postponement, not a solution. It requires a larger cohort to support the last cohort, and a larger one again to support that. Smith has called it a Ponzi scheme, and the description is not hyperbole. It is arithmetic.

There is a second counter that Treasury has not engaged with. Part of what suppresses Australian fertility is the affordability crisis itself: young Australians decide against having children, or against a second or third, because they cannot house the ones they already have or might. Sustained high migration has made housing less affordable, not more. The policy that is supposed to solve the demographic problem is part of what created it.

The empirical case. The Australian Bureau of Statistics recorded that 83 per cent of Australian prisoners in 2024 were Australia-born. The Australian Institute of Criminology reports that migrants as a category show lower victimisation rates than the general population. Both numbers are accurate. Both aggregate in ways that make the aggregate useless.

The Sudanese-born population is about 0.15 per cent of Victoria. It accounts for roughly 1 per cent of alleged offenders in the state – around ten times its population share, concentrated in crimes against the person. The Vietnamese-born population in Victoria, long-established and largely refugee in origin, has a profile closer to its population share. These are different populations, with different profiles, producing different outcomes. A migration policy that refuses to notice the difference cannot respond to it. Denmark noticed. Australia has chosen not to.

This is the case for caution. It is serious. Delay in the current system is itself an integrity failure. The machinery is under-resourced in ways that punish the genuine applicant. The current enforcement emphasis has passed the point of usefulness and become symbolic.

Which is exactly why the right response is not less vetting. It is better vetting: faster, more disciplined, and honest about outcomes by source and by profile, rather than hidden behind an average. The way to respect the people who come through the system is to build a system worthy of their application.

VII. What to do

Three things.

1. Publish the data. By source country. Every year.

The Australian government should follow Denmark's lead. Every year, the Treasurer, the Bureau of Statistics, and the Department of Home Affairs should publish, in one place, what happens to the people we let in, broken down by country of origin. Employment rate. Median income. Welfare receipt. Offending rate. English-language acquisition. Second-generation outcomes once the cohort is old enough to measure.

An average hides the policy question. Denmark has published numbers of this kind for more than a decade, and its migration policy has become responsive to evidence rather than ideology. Australia refuses to publish, and its migration policy has become responsive to neither. If the figures on any particular source country are uncomfortable, that is a reason to confront them, not to leave them unpublished. Alongside this, the Australian National Audit Office should commission the first public performance audit of the migration vetting function since 2012.

The effect is plain. A public, disaggregated outcome table would let the country answer the simplest and most important question a migration program can be asked: who, by source, has contributed, and who has not. It would force every future policy decision to stand against the evidence. And it would end the politically convenient fiction that every migrant cohort produces the same result, when the data shows clearly they do not.

2. Score the sponsor, not just the applicant.

Adopt Singapore's model for employer-sponsored migration. Grade sponsoring firms on their workforce diversity, their local employment record, and their history with existing visa holders. Publish the grades. A firm with a record of exploitation, or with a single-nationality-dominant senior workforce, loses its right to sponsor.

The worst pathologies in our migration system are not caused by individual applicants. They are caused by sponsoring firms that use the system to build captive labour pools, often off their own communities, opaque to inspection and uniquely able to exploit their own people. Grading the firm catches what grading the individual cannot.

3. Ask values questions of everyone.

Every visa stream should include a plain values test. Does the applicant accept the supremacy of Australian law over any religious law? Does the applicant recognise equal standing regardless of sex, sexuality, or creed? Does the applicant reject political violence?

Not by ethnicity. Not by country. Not by religion. By attitude, asked of everyone, with real consequences for dishonesty, and with the machinery to verify the answers against a person's public record where it exists.

Australia is a country of laws and values. Joining it should require assent to them. No serious democracy asks less of its newcomers, and Australia should not be the first.

---

A great migration program rests on a question seriously asked, a question seriously answered, and a record seriously reviewed. We have stopped asking. We have stopped measuring. The country we want to hand on depends on doing both, again.

---

Sources

• Milton Friedman, Mont Pelerin Society lecture on immigration and welfare, 1978.
• Arthur Calwell, Hansard (House of Representatives), 2 August 1945.
• Australian Bureau of Statistics, Overseas Migration, 2024-25 release.
• Department of Home Affairs, Migration Trends 2024-25.
• Australian Security Intelligence Organisation, Annual Report 2023-24, and Mike Burgess, ABC 7.30, September 2024.
• Australian National Audit Office, Administration of the Security Screening of Visa Applicants, Audit Report No. 49 of 2011-12.
• Commonwealth Ombudsman, Administration of Section 501 of the Migration Act 1958, 2016.
• Senate Estimates testimony of Secretary Stephanie Foster, Department of Home Affairs, 2024.
• Direction No. 99 and Direction No. 110 made under section 499 of the Migration Act 1958.
• Abul Rizvi, Pearls & Irritations, commentary on Australian migration policy 2022-26.
• Ministry of Manpower, Republic of Singapore, Complementarity Assessment Framework (COMPASS), effective 1 September 2023.
• Government of Denmark and Wikipedia, Immigration policy of Mette Frederiksen, 2019-26.
• Chatham House, Denmark's immigrants forced out by government policies, 2021.
• Swedish National Council for Crime Prevention (Brottsförebyggande rådet, Brå), Migrants and Crime in Sweden studies, 2005 and 2013, and subsequent peer-reviewed literature in the Journal of International Migration and Integration.
• UK Home Office, Migration Statistics Quarterly Release, 2019-2024.
• Migration Advisory Committee (UK), Annual Report 2024.
• Business & Human Rights Resource Centre, Modern Slavery and the UK Care Sector, March 2025.
• ICM Unlimited for Channel 4 / Policy Exchange, Unsettled Belonging: A Survey of Britain's Muslim Communities, 2016.
• GfK NOP Social Research, Attitudes of British Muslims, 2006.
• Mark Aarons, War Criminals Welcome: Australia, a Sanctuary for Fugitive War Criminals Since 1945, Black Inc., 2001.
• Andrew C. Menzies, Review of Materials Relating to the Entry of Suspected War Criminals into Australia, report to the Attorney-General, 1986.
• Grattan Institute, Rethinking Permanent Skilled Migration, 2021.
• Commonwealth Treasury, 2023 Intergenerational Report.
• Australian Bureau of Statistics, Prisoners in Australia, June 2024.
• Australian Institute of Criminology, Trends and Issues in Crime and Criminal Justice No. 302.
• NZYQ v Minister for Immigration, Citizenship and Multicultural Affairs [2023] HCA 37.
• Pearson v Minister for Home Affairs [2022] FCAFC 203.