Can outsourcing be improved?

A TAL factory in Dongguan, China, in May 2013. The company claims to make one in every six dress shirts sold in the U.S.

The day after the Rana Plaza factory crumbled in suburban Dhaka, Bangladesh, the death toll numbered 225. In the immediate aftermath, a disaster specialist optimistically told the Los Angeles Times, "We were lucky; it could have been much worse." For three weeks volunteers combed through 600 tons of rubble and were occasionally pelted with rocks thrown by outraged crowds. When recovery halted, the final toll was not lucky at all: 1,127 bodies. You'd be hard-pressed to pick a lower point for outsourcing or a better example of the high cost of cheap labor.

The past two decades have provided plenty of reasons to believe that relying on low-wage workers overseas has made multinationals complacent about their safety, beginning with Nike's sweatshops in the early 1990s and continuing up to this past November, when a fire in another Bangladesh factory killed 112 workers, mostly young women, trapped behind locked doors. Some of the companies manufacturing in Bangladesh have rushed forward with promised improvements -- H&M and Inditex (the parent company of Zara) signed an accord to improve laborers' safety and pay in the country. But what would real reform look like? If Rana Plaza represents the worst of outsourcing, is there any model that is best? What can the world learn from companies that get it right?