In 2009 the director of OBI landed in Juba and met with ministers in the South Sudanese parliament. She then traveled to secondary towns, taking photographs of clinics, markets, and schools along the way. After several weeks she flew back to Israel and showed these images on PowerPoint slides to South Sudanese refugees in a community center, informing them that South Sudan had housing, security, free schools, universal healthcare, and income-generating opportunities.

By 2011 several dozen families accepted OBI's assistance to repatriate. After return, most were without reliable shelter, medical care, regular meals, or school. Most notably, they lacked clean water, and had to drink from contaminated rural wells in villages, or streams that flow through mounds of waste in Juba. Some lived off the unreliable charity of distant relatives, or the occasional kind stranger in teashops that dot the streets of South Sudan. While a small number started small businesses, they mostly failed. An unknown number died from illness or ethnic-based violence, and the majority were displaced within two years.

It is widely acknowledged that, if an agent provides a high-risk offer, she must disclose the known risks of this offer. A surgeon must disclose known risks about surgery, a fireworks manufacturer must disclose known risks about fireworks, and the military must disclose known risks of joining the military. But though known risks must be disclosed, it is not clear what risks must be known. OBI did disclose what it knew, but perhaps it ought to have known more, conducting more rigorous research while in South Sudan.

To establish if this is true, we must establish when agents providing high-risk offers have duties to learn the risks of their offers. In some cases, it seems agents have no such duties. If I book a flight to Somalia, my airline needn't tell me the risks of traveling to Somalia. While some agents do have responsibilities to learn about risks, it is not clear when such responsibilities arise.

This ambiguity has serious implications for repatriation, and has been largely overlooked in broader debates on immigration. These debates overwhelmingly focus on when it is wrong to deport or detain immigrants, rather than on when it is wrong to misinform immigrants.