Sweatshops are awful places to work. But they are often less awful than other jobs sweatshop workers could take. And this is the basic argument in defence of sweatshops. When people argue against them, the question we should ask is: “Compared to what?”. Most evidence suggests that sweatshops pay better than the alternatives. It’s hard to collect reliable data in many poor countries, but Ben Powell and David Skarbek’s 2006 paper “Sweatshops and Third World Living Standards” uses wage data given by anti-sweatshop campaigners­ to estimate wages for sweatshop workers in ten countries compared to average National Income. This, if anything, should underestimate sweatshop workers’ earnings.

Again, it’s difficult to know how many hours the average sweatshop worker does every week, but most anti-sweatshop campaigners suggest that it is more than 70 hours per week. The results should be taken with a pinch of salt, but Powell and Skarbek found that sweatshop wages exceed average income in between eight and ten out of ten countries surveyed, depending on how many hours were worked.

In nine out of ten countries, “working ten-hour days in the apparel industry lifts employees above (and often far above) the $2 per day threshold.” And “in half of the countries it results in earning more than three times the national average”! (Powell's defence of sweatshops, here, is excellent. His book on the topic is self-recommending.)

Critics of sweatshops point to the 1,000+ people killed and 2,500+ people injured by the collapse of the Rana Plaza sweatshop in Bangladesh in 2013. This was indeed grotesque, and evidence of the poor conditions that many sweatshop workers have to work in.

But what is their next-best alternative? Subsistence farming still dominates many of the countries that sweatshops operate in – in Vietnam, 59% of workers are self-employed in farming; 1.5% work for businesses owned partially or fully by foreign firms. And farming – particularly subsistence farming – is one of the most dangerous occupations in the world.

The International Labour Organisation estimates that agricultural workers suffer 250 million accidents every year, and say that in some countries the fatal accident rate is twice as common in agriculture as in other industries. “Out of a total of 335,000 fatal workplace accidents worldwide,” say the ILO, “there are some 170,000 deaths among agricultural workers.” As horrendous as the Rana Plaza incident was, anti-sweatshop campaigners have not shown that sweatshops are more dangerous than sweatshop workers’ next-best alternative.

Sweatshops seem to have good impacts on women in particular. A study by researchers at the Universities of Washington and Yalethat I blogged about last year looked at different villages in Bangladesh – some close to sweatshops, some not.

In the villages close to sweatshops, girls were substantially less likely to get pregnant or be married off (28% and 29% respectively, and this effect was strongest among 12-18 year olds) and girls’ school enrolment rates were 38.6% higher. The authors say that these effects were likely due to a combination of wealth effects (richer families need to marry off their daughters less early, and can afford to send their daughters to school for longer) and the fact that garment factory jobs reward skills, increasing the value of education.

And what do workers themselves think of sweatshops, given not just wages but other non-monetary compensation as well? Using field interviews with thirty-one sweatshop workers in El Salvador, David, Emily, Brian and Erin Skarbek found that “workers perceive factory employment to provide more desirable compensation along several margins.”

This is not to condemn all work done ‘against’ sweatshops. Using data from Indonesia, the World Bank's Ann Harrison and Jason Scorse found that 1990s campaigns to improve conditions for sweatshop workers in the developing world seem to have led to real wage increases without significant unemployment effects, though some smaller factories did close.

The lesson here may be that work that focuses on improving wages and conditions for sweatshop workers, not closing down sweatshops and trying to wash our hands altogether, may be the best approach. Persuading consumers to continue buying things from sweatshops, but to pay a higher price to give those workers a better wage, might be a decent way of essentially 'bundling' a charitable donation into a normal purchase. Unfortunately, most campaigns in Britain seem to be straightforwardly anti-sweatshop.

And even the most noble-seeming campaigns can backfire. UNICEF argues that early 1990s campaigns to reduce child labour in Bangladesh’s formal economy led to children looking for income in much worse places: stone-crushing, street hustling, and prostitution.

It is understandable that anti-poverty campaigners find sweatshops appalling, and work done to improve conditions in sweatshops might be valuable, but too often people forget that blunt campaigns against sweatshops probably end up hurting people. Instead, people should use the awfulness of sweatshops – and even greater awfulness of other jobs – as proof that we need to do more, much more, to give better options to poor people in other countries.

One option might be guest worker programmes, targeted at people from the poorest countries in the world, to allow them to come and work in the developed world so that they can send more money back home for investment. And lower trade barriers to goods from poor countries would help them grow, too.

Sweatshops are particularly horrifying because they make us feel complicit in the suffering of the poor. They are not a good option, but they are the least bad option currently available to many people. Washing our hands of the situation and just closing the sweatshops would make their workers worse off, potentially much worse off. If we want to help people, we should give them new options, not take away existing ones.