Why should we be concerned about the XO and crime? Probably because the extent and pervasiveness of crime in the developing world is something not always understood from outside.
I am Eduardo Villanueva Mansilla and Lima, Peru, my home town, is not precisely the crime capital of the world; there are worst places around, like San Salvador or Rio de Janeiro; even Caracas has significant problems. Problem is the penetration of criminal activities into everyday life.
You see, from time to time, DSL services stops in the well-off neighbourhoods where it's relatively common. The usual reason? Cables are stolen. Even fiber-optics cables, though they are worthless, unlike copper ones that can be re-used. To steal a bunch of copper wires, robbers are willing to grab all the cables around. The actual benefit is very small, a couple of USD for 100 meters of cable. But the service disruptions occur anyway.
Everybody knows where the cables are stolen, where the cables are available, and how little they are worth. From time to time, a robber gets caught, and released by noon because the monetary value of the stolen goods is too small to merit prosecution. So he returns to the same and steals cables, again and again, all around town, shutting down DSL services, phone services, whatever.
Something similar happens all the time, of course at different intensities and for different goods, in most of the developing world. In many cases, violence is part of the equation; but it is not necessary for it to occur for crime to be so annoying: it's the uselessness of the petty criminality that drives everybody crazy.
Crime is a mixture of vandalism and actual stealing of anything that can be re-sold, eventually to the original owners, like the many times somebody gets his car's windshield stolen just to see it next day at the flea market where all the stolen goods are sold, and paying to recover it since there are no alternatives besides a extremely expensive original one.
This points towards the actual problem. It is not just crime, is the complete absence of any mechanism to make crime worthless. If the flea market was raided and all stolen goods recovered, there will be no actual incentive to steal wires or windshields or even cell phones.
Problem is, the flea market is just one of many, this particular one covers four or five city blocks, and a full ecology and economy lives off it. To raid it will mean to have an angry mob of a few thousand fighting the police. It is easier for everyone to just live and let live. The cost of doing business.
There are technical measures, like Bitfrost, to avoid XO computers being stolen, or at least to locate them. For them to work, some kind of institutional mechanisms would have to be in place, allowing the actual inactivation of the machines, and the black market should be fought by law enforcement.
If one kid from a group of thirty gets her Children's Machine XO stolen, she shouldn't have to request a replacement nor buy it back from the black market: she should get help from the police to recover it. It won't happen. Not at the scale necessary for it to be the exception instead of the rule.
Even flooding the crime ridden areas with OLPC XOs will only mean that stealing them will become a shakedown instead of an actual "reselling". It is also possible that at least some teachers will get their piece of the action, putting pressure on the students to get the X0 they need for they to pass to the next school year, or whatever.
Criminal gangs raiding small, rural towns for XOs to be sold a below-market prices (10 USD?) in big cities to compensate for the losses due to robberies. A lot of other awful scenarios.
Again, it is the nature of crime in our societies that make the situation difficult: the combination of institutional inability to confront crime, the large number of people involved in crime at many different levels, the common-place nature of stolen goods, the inevitability of XOs becoming "required material" for passing to one grade to another. Too many variables. Too many interests. Too little law enforcement capacities.
This is, obviously, an extreme scenario. But to disregard completely the significance of crime in developing nations while planning for the implementation of the XO may be a severe mistake.