Cities of Refuge

When I first met Paul Roemer a few years ago a few years ago he was promoting an idea he called “Charter cities.”

The basic idea is that, under contemporary economic conditions – in particular the astonishingly low cost of transporting freight by water – economic activity doesn’t require much in the way of natural or even human resources: just a piece of land with access to ocean transportation, a little basic urban planning, the rule of law, and the absence of intrusive and kleptocratic government.

Alas, those last two are sufficiently rare in the worst-off countries that hundreds of millions of people, deprived of industrial opportunity where they are, want to leave and go somewhere else. The problem is that “somewhere else” mostly doesn’t want to take them in.

Romer’s proposed solution flows directly from this analysis: find an empty space with the requisite transportation access, get the national government out of the way, and build a new city, attracting economic migrants from around the world. At first the charter city would be governed by an international board, with housing and industrial plant built by private actors and infrastructure and public-services financed by ground leases: in effect a Henry George single tax.

Over time, the charter would transition to democratic self-rule, independent of the country from which it had been carved out. By the same token, the migrants to the charter city and their descendants would be its citizens, not citizens of the host country.

Of course there were and are countless details to be worked out: Romer called me to talk about how to police the place. But the big problem – the insuperable problem I thought at the time – was convincing one of the very governments whose incompetence and corruption made the proposal necessary to create such a new entity.

For a while, it looked as if Honduras might be willing to play, but instead the government decided to make the governing body a self-perpetuating oligarchy; the transition to democracy was dropped from the proposal. That led Romer to walk away from the Honduran situation, remarking that he wouldn’t be involved in setting up a charter city unless he’d be happy for his grandchildren to live there.

As a result, I had pretty much written off the idea as one more thing that would be nice to think about but wasn’t actually going to happen. Too many of the people who would need to support the idea for it to go into practice were likely to see it as one more libertarian dystopia in the making.  And there’s always the danger that some magician will solemnly pronounce the incantation “Colonialism” to make clear thought impossible.

That discouragement didn’t keep me from joining Romer in his new enterprise, the Marron Institute; there’s plenty of work to do managing actual cities even if we never get to build an imaginary one.

But Romer, being persistent as well as clever, has now found a more pressing problem to fit his solution. The world now has roughly 50 million refugees, and on current form they are likely to spend an average of twelve years living in refugee camps where local rules forbid them to engage in manufacturing. They lead very tough lives, and constitute a burden on the countries where the camps are built and a financial drain on world institutions.

So Charter Cities are no more; the new proposal is for Cities of Refuge. The design is roughly the same; all that’s needed, to build a city designed to reach a population of 10 million, is roughly 10 miles of coastline. Perhaps surprisingly, there is a fair amount of virtually uninhabited coastal real estate in the world; for example, the coast of Namibia is unpopulated because it doesn’t have water, while the Mediterranean coast on both sides of Libya is unpopulated because ISIS and other armed actors make it unsafe.

So, the idea goes, let one of the UN agencies buy a 99-year or 999-year lease on an appropriate piece of territory whose autonomy would be guaranteed by international agreement, and throw the new place open to refugees. The money now being spent on building camps and providing minimal food and shelter to the people in them could be spent instead on the initial lease payment, on a security force if necessary - it’s not very hard to defend a small bit of territory against militias - and on things like artificial ports and desalinization plants. And in terms of “human capital” - education, skill, and experience - refugees tend to be much richer than typical economic migrants.

I’m less confident than Romer is that we can build a Singapore anywhere we want to; the economics may work, but the politics and the sociology are going to be tough. Republican government and the rule of law aren’t easy tricks to pull off, and civil society isn’t built as quickly as a factory.  (Just think about what sort of religious institutions you’d want the city to have, and the difficulty of creating such institutions without turning them into a state church.) Everything gets harder when the population doesn’t share a common culture.

Still, even a fairly sketchy city beats a refugee camp hands-down.

Getting to actually do it someplace still seems like something of a long shot, even with Europe in a panic about how to handle the flow of refugees from the Middle East.  But now I think of it as a long shot well worth playing. If it works, the payoff is almost incalculably great.