In December 2018, delegates from more than 150 countries gathered in Marrakech to adopt the first comprehensive global framework on migration. The Global Compact for Safe, Orderly and Regular Migration was explicitly non-binding. It reaffirmed state sovereignty in its opening pages. It created no new body with enforcement power. And yet it triggered government crises, street protests, and a wave of withdrawals from countries that had helped negotiate it.

That reaction reveals something important. Roughly 280 million people live outside their country of birth, and migration shapes labour markets, remittance flows, demographics, and security in nearly every state. Few issues are more obviously transnational. Yet migration remains one of the least governed areas of international life. There is no World Trade Organization for people, and no binding migration treaty that major destination states have accepted.

The real puzzle is not why migration governance is difficult. It is why governance stays so thin while the need for coordination keeps growing. The answer lies in three things: a historical exception, a sovereignty reflex, and a newer style of soft diplomacy that has struggled to deliver results.

The 1951 Convention Relating to the Status of Refugees came out of a specific moment. Postwar Europe held millions of displaced people, and the memory of the 1930s was still fresh. At the 1938 Évian Conference, states had expressed sympathy for Jewish refugees and then accepted almost none. The Convention defined a refugee as someone with a well-founded fear of persecution because of race, religion, nationality, social group, or political opinion.

Its core is non-refoulement, set out in Article 33. States may not return a refugee to a territory where their life or freedom would be threatened. This is a real constraint on sovereignty: a binding legal duty, supervised by a dedicated agency, UNHCR, and accepted by around 150 states. The 1967 Protocol removed the Convention's original time and geographic limits, which turned a European arrangement into a global one.

Nothing comparable exists for other migrants. The 1990 Migrant Workers Convention took thirteen years to enter into force. It has been ratified mostly by countries of origin, and no major Western destination state has joined. The International Organization for Migration spent most of its history outside the UN as a service provider to governments. It became a related organization only in 2016, and it still has no formal protection mandate.

The result is a binary system. A person fleeing persecution has rights. A person fleeing drought, economic collapse, or widespread gang violence often falls into a gap, a situation the scholar Alexander Betts calls 'survival migration.' Mixed flows across the Mediterranean or the Darién Gap include both groups. Because asylum is the only strong legal door, it comes under pressure, and states respond by narrowing it through offshore processing and safe-third-country deals.

Takeaway

The refugee regime is strong precisely because it was framed as an exception, a humanitarian carve-out from state control. That same framing makes it hard to extend protection to anyone else without seeming to threaten the exception itself.

Why do states guard migration so closely? Deciding who enters is ultimately deciding who belongs: future workers, taxpayers, voters, and citizens. Trade in goods changes prices. Migration changes the makeup of the political community itself. The philosopher Michael Walzer argued that control over admission sits at the heart of communal self-determination. Most governments, whatever their politics, quietly agree.

There is also an asymmetry of interests. Trade liberalization runs on reciprocity: I open my market, you open yours. Migration flows tend to run in one direction. Origin countries want access for their workers and the remittances they send home, which now exceed 650 billion dollars a year to low- and middle-income countries. Destination states hold the leverage and see little benefit in binding commitments. This explains the lopsided ratification record of the Migrant Workers Convention.

Domestic politics adds to the resistance. Migration is highly visible and electorally potent, and a perceived loss of control carries heavy political costs. The Brexit slogan 'take back control' captured a feeling that extends far beyond Britain. International agreements become easy targets when they can be framed as foreign interference. Even liberal governments prefer flexibility, so they can adjust admissions to economic cycles and public mood.

Yet states do cooperate on migration, just selectively. Schengen, EU free movement, the ECOWAS mobility protocol, bilateral labour agreements, readmission deals, and regional forums like the Puebla and Budapest Processes all exist. These are the government networks Anne-Marie Slaughter describes: interior ministries, border agencies, and consular officials sharing data and practices. Cooperation flourishes around control, enforcement, and selective access. It is rights-based multilateralism that stalls.

Takeaway

States do not refuse to cooperate on migration; they refuse to cooperate in ways that limit their discretion. Look at where cooperation thrives and you will see what sovereignty is actually protecting.

The compacts grew out of crisis. In 2015, more than a million people reached Europe, many by sea. In September 2016, the UN General Assembly adopted the New York Declaration for Refugees and Migrants, which committed states to negotiate two separate compacts: one on refugees, led by UNHCR, and one on safe, orderly and regular migration, negotiated by states with Mexico and Switzerland as co-facilitators. The decision to separate them reinforced the old binary.

The migration compact set out 23 objectives, from better data and legal pathways to reducing the drivers of displacement and cooperating on returns. It was deliberately non-binding and affirmed each state's right to set its own migration policy. Even so, the United States withdrew from negotiations in December 2017. In late 2018, Hungary, Austria, Australia, Poland, the Czech Republic, and others pulled back. In Belgium, the dispute brought down the governing coalition. The final General Assembly vote was 152 in favour, 5 against, and 12 abstentions.

The refugee compact passed more easily, with 181 votes in favour. Its focus is responsibility-sharing, because most refugees live in neighbouring countries that are often poor. The Global Refugee Forum, held in 2019 and 2023, gathers pledges of money, resettlement places, and policy changes. Those pledges remain uneven, and resettlement still reaches only a tiny share of refugees in any given year.

Implementation relies on soft tools. The International Migration Review Forum meets every four years, and a UN Network on Migration, coordinated by IOM, supports states that volunteer to take part. Funding is modest and enforcement is absent. Critics call the compacts too weak to matter. Supporters point to early environmental frameworks, where soft commitments built shared vocabulary, data, and peer pressure before anything harder became possible.

Takeaway

The fierce backlash against a non-binding document shows that in migration, symbolism is itself a form of power. The compacts may change little in law, but they contest who gets to define the problem.

Judged honestly, global migration governance is a patchwork. It combines a robust but strained refugee regime, a thin framework for everyone else, and dense networks of cooperation focused mainly on enforcement.

Progress is unlikely to come from a grand treaty. More promising paths build on shared practical interests: skills partnerships that train workers in origin countries for jobs abroad, portable pensions, regional mobility arrangements, and better data. Regional instruments such as Africa's 1969 OAU Convention and Latin America's Cartagena Declaration already define refugees more broadly. They offer models for protecting people displaced by climate disruption and widespread violence.

The deeper lesson is that lasting cooperation grows where states see migration as a shared interest rather than an imposed obligation. The open question is whether that can be built before the next crisis forces it.