Above: Migrants rescued by the ship MS Aquarius arrive in Valencia, Spain, aboard an Italian military vessel on June 17.  (Photo by Pablo Blazquez Dominguez/Getty Images)

629 asylum seekers were stranded on the Mediterranean Sea last week when Italy refused to allow their disembarkation. The group, which included 134 children and seven pregnant women, was picked up from inflatable rubber boats on the Mediterranean on June 9 by the NGO rescue vessel, MS Aquarius.

The Italian government attempted to shift responsibility to Malta, which also refused to admit the rescuees. With conditions on the overcrowded rescue vessel deteriorating, a potential humanitarian disaster was averted only after Spain’s Prime Minister, Pedro Sanchez, intervened to accept responsibility.

Italy’s actions must be condemned. The longstanding principle that a shipmaster has a duty to rescue persons in distress without regard for their nationality, status, or circumstances is only viable so long as states honor their duty to enable the speedy disembarkation of those rescued – a duty that Italy (and perhaps also Malta) breached in this case.

Emmanuel Macron denounced Italy’s actions as cynical and irresponsible. But the Italian Prime Minister, Guiseppe Conte, was to some extent justified in calling out the hypocrisy of these comments.

Italy and Greece have shouldered far more than their fair share of refugees seeking protection in Europe. The European Union’s ‘Dublin Regulation’ unfairly places sole responsibility for affording protection on the first country in which a refugee arrives – meaning that states bordering the Mediterranean face heavy burdens while others states can wash their hands of what is in principle a common responsibility.

Nor is it an answer to call for temporary, ad hoc redistribution of the kind that the EU agreed to during the height of arrivals in 2015, and which the UN’s refugee agency (UNHCR) now proposes as the global standard under its draft “global compact.” Once states know where refugees have arrived there is little incentive for them to pitch-in and lessen the burdens of others. Much less is there an incentive for frontline states to keep doors open to those in need if there is a real risk that sharing will not happen.

Instead, an insurance-based approach to protection is needed, under which, fair shares are defined in advance and implemented in an efficient and dependable way. Such reform will be dramatically more viable if Europe does not simply attempt to redistribute refugees within the EU. It should instead partner with safe states in regions of origin — and ultimately with states around the globe — to collectivize the provision of asylum.

The basic premise would be that asylum seekers stand on an equal footing no matter where they apply for protection. There would no longer be any incentive for asylum seekers to attempt longer and riskier journeys to reach wealthier destinations where they can maximise their economic and social outcomes. Nor would there be any incentive for wealthy nations to put up barriers preventing asylum seekers from accessing their territories.

Under a collective approach, there would be one unified refugee processing system. A person could apply for asylum from any country. If found to be in need of protection, they would then go into a centralised pool, to be provided protection for the duration of risk in a safe country. This will normally be in their region of origin in order to maximize the functional and cultural compatibility needed quickly to get back on their feet. Preference matching algorithms can be used to maximize refugee autonomy. Host countries would be guaranteed both the resources needed to provide asylum and start-up grants to enable refugees and host communities to thrive.

Within 5 years, history suggests that at least half of refugees would be able either safely to go home or to integrate locally. Our proposal builds on that reality to renew asylum capacity. But for refugees unable to go home or to stay in the first asylum state by the 5-year switch point, the revamped protection system we advocate would guarantee resettlement outside the region. The number of resettlement spots needed would be roughly the same as the number of asylum cases now processed each year in wealthier countries, allowing Europe and other OECD states to swap their present chaotic and absurdly expensive asylum systems for a more managed, orderly role.

Under this system, a refugee in Africa or the Middle East would have the same chance of resettlement in Europe as one who risks his or her life to reach Europe’s shores. And refugee smugglers would be out of business because they would have no immigration result to sell.

It would end the disputes of the sort we saw in the Mediterranean this week. States could grant asylum seekers access to their territory, safe in the knowledge that no immigration consequence would follow from that admission.

The current approach to refugee protection is clearly broken. Urgent action is required to end inter-state squabbling over responsibility for affording protection. Until then, refugees will face more deflections, more deterrence and more precarity.