ABSTRACT
A few years ago I received a phone call asking me to give money to buy the freedom of a slave (in the Sudan). The cost, and sacrifice to me, would be extremely modest – no more than $50, or dinner for two. It seemed an incredibly compelling ask. Nonetheless I was reluctant to give, fearing that I would just be reducing the supply of slaves and thereby driving up the “price.” That seemed to me to create the potential to draw more people into the slave trade and thereby increase the number of slaves. My fear was that freeing a slave might result in more than one person being newly enslaved and so make things worse off than they were. With the help of some economists, I found out my intuitions were misplaced. Redeeming the freedom of one slave (under most circumstances) does no harm and likely does at least some good. It turns out if I buy the freedom of 10 slaves, no more than 10 people will be newly enslaved and likely less than 10 will be enslaved. So, all other things being equal, net-net, I do more good than harm by buying their freedom (for details, see Karlan and Krueger 2007). There is a complication in this analysis. It assumes that the costs of slavery can be shared so that two people enslaved for 6 months each is no worse than one person enslaved for 1 year. But that may not be the case. Suppose systematic rape is an accompaniment of the act of enslavement. But suppose we can be sure this complication does not hold. Then ought I not to give? “No” Peter Singer might say, “you can bring about a greater good by giving your money to feed starving people.” (In fact that is just what Singer did say to me about this issue in conversation.)
