RWANDA has developed a strategic plan for them. Myanmar is embracing them as it opens up.

Countries that have long been fans, from China to the United Arab Emirates, are doubling down.

India’s plans in the area are “revolutionary” and could add 2% to its GDP, says a minister.

Special economic zones (SEZs) are all the rage among governments hoping to pep up their trade and investment numbers. Such havens are appearing even within havens: the Cayman Islands has a new one. “Any country that didn’t have [an SEZ] ten years ago either does now or seems to be planning one,” says Thomas Farole of the World Bank.

Studying history may give eager trade ministers pause. SEZs—enclaves in which exporters and other investors receive tax, tariff and regulatory incentives—create distortions within economies. Other costs include required infrastructure investment and forgone tax revenues.

The hope is that these are outweighed by the boost to jobs and trade. In reality, many SEZs fail.

Performance data are elusive because the effects of zones are hard to disentangle from other economic forces. But anecdotal evidence suggests they fall into three broad categories: a few runaway successes, a larger number that come out marginally positive in cost-benefit assessments, and a long tail of failed zones that either never got going, were poorly run, or where investors gladly took tax breaks without producing substantial employment or export earnings.

SEZs have a long pedigree: the first free-trade zones were in ancient Phoenicia. The first modern one was set up at Shannon airport in Ireland in 1959, but the idea took off in the 1980s after China embraced them. There are now more than 4,000 SEZs (see chart). A study conducted in 2008 estimated that 68m people worked in them. They come in many forms, from basic “export processing zones” to “charter cities”, urban zones that set their own regulations in all sorts of areas that affect business.  
.

The biggest success story is China, whose decision in 1980 to create a zone in Shenzhen transformed the city (pictured) into an export powerhouse. Dozens of SEZs have since popped up across the country. In March, Xi Jinping, the president, urged a faster pace of roll-outs.

Other successes include the United Arab Emirates, South Korea and Malaysia. The Philippines has won praise for its “PEZA” zones, which offer a streamlined permit process for foreign investors, says Shang-Jin Wei of the Asian Development Bank.

Most economists agree that SEZs catalysed liberalisation in China, which used them to test reforms that were seen as too hard to unveil nationwide. In the Dominican Republic they helped create a sizeable manufacturing sector in an economy previously reliant on agriculture.

The overall impact of SEZs on trade is poorly understood. A paper published in 2014 by economists at Paris-Dauphine University found that, for a given level of tariff protection, SEZs increase exports for the countries they are in and for other countries that provide intermediate goods or components.

This helps explain why the World Trade Organisation generally tolerates SEZs, even though many breach its subsidy rules. However, the paper also concluded that zones sometimes give countries an excuse to retain protectionist barriers around the rest of the economy.

More prosaic problems pop up, too. Bureaucracy can be excessive, and the bureaucrats underfunded—sometimes at the same time. Too little is often spent on railways, roads and ports to link the zone to the rest of the world. Many African SEZs have struggled for such reasons. One in Senegal flopped because of a combination of excessive bureaucracy, high electricity costs and its distance from a good port.

Developers have withdrawn from 61 of the 139 approved SEZs in the Indian state of Maharashtra because of capricious policymaking, a murky screening process and concern over economic prospects. One survey found that firms sometimes had to deal with 15 different agencies to do business in an Indian zone. Violent protests by locals over land acquisition for zones have also deterred investors.

Moreover, governments sometimes embrace SEZs for the wrong reason: to win praise for reform (and votes) without having to risk full liberalisation. Partial liberalisation can also be a way to preserve some of the rents earned elsewhere by shielding businesses from competition.

Some officials see zones as vehicles for graft. In 2005 some 60% of firms in Indian SEZs reported having to make “irregular” payments to zone authorities. Last month Ukraine’s prime minister said he opposed SEZs because of corruption. SEZs in Nigeria were firmly resisted by the customs agency, which did not want to lose its clout. Another concern is the use of zones to launder money, by inflating export values.

The SEZ concept appears to have natural limits, too. What works in manufacturing may not work in other sectors. The Shanghai Free Trade Zone, launched in 2013 and focused on finance, has been disappointing. Economists fret that it is impossible to tinker within the zone with China’s capital controls, for instance, without the effects spilling over to the rest of the economy. Perhaps as a result, the authorities have been cautious: in a recent survey, three-quarters of American firms in Shanghai said the zone offered them no benefits.

That hasn’t stopped China approving plans for more financial SEZs. The government is also promoting zones abroad: it is helping six African countries to set some up. Although its are state-run, ever more SEZs are likely to be privately owned and operated. The Philippines already has more than ten times as many private zones as public ones. This shift may go further, if privately run charter cities and other so-called “special governance zones” gain traction. The idea is to create enclaves that write their own rules in all business matters, from labour regulation to anti-corruption codes—“to look at laws as services that companies demand”, says Lotta Moberg of George Mason University.

Such ventures will provide competition more effectively than zones focused on fiscal incentives, says Shanker Singham, founder of Enterprise Cities.

Mr Singham is in talks to develop sites in the Dominican Republic, Colombia, Morocco, Bosnia, India and Oman. But these are mostly at an early stage. The most advanced charter-city project, backed by a group of American libertarians, is in Honduras. But it has yet to start and is already controversial: many Hondurans worry that it will operate as a state within a state, milked by business interests. In most countries, such parastatal ventures are likely to encounter political difficulties.

Whether or not such freewheeling zones catch on, expect more experiments. South Korea and Thailand are developing eco-industrial parks. Others are considering SEZs for refugee populations.

For better or worse, the number of zones could top 5,000 before long.