CHRISTIANA ESSIE SAGAY

“Almost all nations observe almost all principles of international law and almost all of their obligations almost all the time.”

— Louis Henkin, How Nations Behave

I keep returning to Henkin. Not because the line is comforting, but because the qualifiers are doing such heavy lifting. Almost all. Almost all. Almost all of the time. The space between those almosts is where the lived experience of migrant workers, and, increasingly, the people who depend on their care, quietly accumulates.

Last week, between 5 and 8 May 2026, the world’s Member States convened at UN Headquarters in New York for the Second International Migration Review Forum (IMRF), the quadrennial stock take of the Global Compact for Safe, Orderly and Regular Migration (GCM). Speeches were made. A Progress Declaration was negotiated and adopted. Diplomats committed, again, to facilitate “access to social protection for migrant workers and their families.” Trade unions and civil society pressed Member States on family reunification and decent work.

And while New York deliberated, a quieter, and far more consequential set of rules continued to bite in London.

Illustration by Pragati Choudhari on Unsplash   

Transnational care, the practices, labour, and relationships of care that stretch across national borders, is the seam where the contradictions in global migration governance show themselves most plainly. The migrant carer in Manchester is herself part of a family. She is a daughter, mother, sister, friend somewhere else. The care she gives here is shadowed by the care she cannot give there. And the legal regime that brought her here has, in the two and a half years since the Home Office’s December 2023 announcement of a forthcoming dependant ban, become considerably less hospitable to that fact.

Here, I ask questions I find myself unable to set down: what does it mean for a State to participate in the global review of a migration compact while simultaneously narrowing the very pathways the Compact urges it to expand?

For context, let’s examine what changed between December 2023 and May 2026. A brief chronology helps show how the story evolves

1. 11 March 2024. The dependent ban took effect. New care workers and senior care workers (SOC 6135/6136) could no longer sponsor partners or children to join them in the UK.
2. 22 July 2025. Overseas recruitment of care workers closed entirely. The route through which approximately 100,000 carers entered the UK between 2022 and 2024 was simply turned off. In-country switching is permitted only until 22 July 2028.
3. November 2025. The Home Office published "A Fairer Pathway to Settlement", a consultation paper proposing to extend the qualifying period for Indefinite Leave to Remain (ILR) from 5 years to a 10-year baseline, and to 15 years for low- and medium-skilled workers, with social care workers expressly named in that category. The consultation closed on 12 February 2026 with over 200,000 responses. The Home Secretary has signalled that the new rules may apply retrospectively to those already in the UK who have not yet secured ILR, meaning the existing cohort of care workers could find the goalposts moved on them.
4. 8 January 2026. English language requirements for new applicants and their dependants rose from B1 to B2 under the Common European Framework of Reference.
5. March 2026. The UK introduced a temporary ban on visa applications from abroad for nationals of Afghanistan, Cameroon, Myanmar, and Sudan.

The cumulative effect is consequential. In roughly twenty-six months, the route has been closed to new overseas applicants, family reunification prohibited for those still arriving, settlement deferred, and the language bar raised mid-journey. Approximately 120,000 existing care workers remain in the UK on a route whose terms have been rewritten around them.

The UK’s argument, some would say, deserves engagement. It would be wrong to pretend the UK has acted in a policy vacuum. Net migration to the UK reached historic highs in 2022 and 2023. The care worker route, opened in 2022 to relieve acute social-care staffing shortages, expanded faster than anyone, including the sector itself, anticipated. Sponsor licence abuse, exploitative employers, and outright sham sponsorships became a real and well-documented problem. Successive governments faced electoral pressure to reduce headline migration figures and to clean up a route in which abuses had proliferated. “Do something about migration” is, in democratic terms, a legitimate demand.

The successive restrictions can therefore be understood as a response to a genuine policy dilemma: how to keep the social care sector staffed without sustaining a visa route plagued by exploitation, and without contributing to net migration figures the electorate has signalled it finds excessive.

That is the argument as I understand it. Some may say, well, this is not a bad-faith argument. Others may respond to that assertion by saying it is incomplete, because it treats the UK’s international commitments as background scenery, when the IMRF, in which the UK has just participated, suggests they should be in the foreground. 

But let’s turn to the GCM and the coherence question. When the GCM was adopted in 2018 and endorsed by UNGA Resolution 73/195, the United Kingdom voted in favour. It signed up to 23 Objectives. The ones most directly relevant to the current regime include Objective 5 (enhance availability and flexibility of pathways for regular migration), Objective 6 (facilitate fair recruitment and decent work), and Objective 16 (empower migrants and societies toward inclusion and social cohesion).

The IMRF 2026 Progress Declaration goes further. Member States jointly recognised “the continued concentration of many migrants, including migrant domestic and care workers, most of whom are women, in the informal economy, low-paid jobs that are excluded from labour law protections”, and committed to “facilitate access to social protection for migrant workers and their families.” 

The GCM is, of course, soft law. Its paragraph 7 expressly disclaims binding effect, and States retain “sovereign right to determine their national migration policy.” This is a real and proper feature of the Compact, not a flaw. But the soft-law character of the GCM does not entirely insulate a State from a coherence question. To negotiate, vote for, submit a Voluntary GCM Report to, and actively participate in the review of a Compact, while domestic policy moves in the opposite direction on several of its central Objectives, is at minimum a question of multilateral good faith. And good faith in multilateral cooperation, the kind that the International Court of Justice in Nicaragua treated as meaningful even where binding rules were absent, is the soil in which collective action on migration grows. Erode it, and the Compact itself becomes harder to defend the next time someone questions its purpose.

This is the Henkin point, restated. “Almost all nations observe almost all principles.” But which obligations get softened, and whose lives those almost are written on, is itself a question worth asking out loud.

Beyond the GCM sits a body of harder law, and here the picture sharpens, particularly once you ask the gender question directly. Who, in practice, is the UK care worker? Reports show that, 79% of the adult social care workforce in England is female. And visa data shows that of the 71,000 care worker visas granted in 2023–24, more than three-quarters went to nationals of just five countries: India, Nigeria, Zimbabwe, Ghana, and Bangladesh. She is, in other words, very often a Black African or South Asian woman, frequently the primary economic and caregiving support for a household stretched across two countries at once.

That description, racialized women migrant workers, in low-paid jobs, tied to a single employer, facing barriers to settlement, is not a description the international community is encountering for the first time. It is precisely the cohort that the Convention on the Elimination of All Forms of Discrimination Against Women (CEDAW), which the UK ratified in 1986, has long been asked to address — and in 2008, the CEDAW Committee did so directly. General Recommendation No. 26 on Women Migrant Workers identifies women migrant workers “in low-paid jobs… who may be at high risk of abuse and discrimination and who may never acquire eligibility for permanent stay or citizenship,” and calls on destination States to ensure their residency does not depend on the goodwill of a single employer, and to enable family reunification on terms that do not disadvantage women. That cohort, tied to a single sponsor, denied dependants on arrival, and now facing a possible 15-year wait for settlement, is, without overstating it, the situation GR 26 was drafted to address.

The family-life dimension reinforces this. Article 10(1) of the ICESCR, to which the UK is a State party, requires that “the widest possible protection and assistance should be accorded to the family.” And Article 8 of the European Convention on Human Rights, incorporated into UK domestic law via the Human Rights Act 1998, protects the right to respect for private and family life. Strasbourg has been clear, including in Jeunesse v. Netherlands, that the margin of appreciation in immigration matters, while genuinely wide, is not unlimited; and that categorical measures benefit from individualised proportionality assessment.

I am alive to the strongest counter-argument: that the dependant ban applies only prospectively, that those already in-country with dependants in place are unaffected, and that the State retains a legitimate interest in regulating who enters. All true. The interesting question is whether a categorical prospective ban on family unity for an occupational category selected precisely for its indispensability to the social care economy survives the proportionality test as comfortably as the policy assumes. Some lawyers will disagree. That the question is genuinely contestable, however, is itself the point.

One footnote worth flagging: the UK has never ratified ILO Convention No. 97 or No. 143, the two principal migrant worker conventions, nor the ICRMW. That is the UK’s sovereign right. But the absence of those ratifications shapes the doctrinal terrain on which the current policy choices are being made, and the IMRF 2026 outcome renews the call on all States to reconsider.

Going back to the title of this blog, so, what is the obligation, really? International law, in its current form, is not going to compel the UK to reopen the care worker route or reinstate dependent visas. There is no successor to Nicaragua coming on care worker mobility and probably shouldn’t be.         

But that was never the only question. The other question is the one Henkin's qualifiers gesture toward, about the almost-observed obligations, the bracketed principles, the partial compliance, and it becomes most visible when held against the workforce this blog has been describing: a cohort that is overwhelmingly female, often racialized, often the primary caregivers of children growing up without them. On the other side of the bargain are those in need of care, whose well-being depends on the continued willingness of this workforce to migrate under terms that grow steadily more austere.

The IMRF closed last Friday with the adoption of its Progress Declaration. The next four-year cycle begins now. Member States will produce their next round of Voluntary GCM Reports, and the United Kingdom will, presumably, prepare its own. That report will be a small but useful moment of accountability, a chance to describe, in narrative form, how the regime as it stands gives effect to the Objectives the UK signed up to.

I hope it is a good-faith document. 

I hope the gap it has to explain is one the UK takes seriously enough to start, in time, to close.

In the meantime, naming the gap precisely, in the language of obligation rather than the language of grievance is the work.

The IMRF having now closed, the next four years will demand it of all of us who take the Compact seriously.