The government has decided to commission a new, privately run prison. As Eric has previously discussed, there is a fairly canonical paper on the topic by some Harvard economists, which concludes that:
…a plausible theoretical case can be made against prison privatization. This case is weakened if competition for inmates can be made effective, but strengthened by the relevance of political activism by private contractors. One instance in which the case against prison privatization is stronger is maximum security prisons, where the prevention of violence by prisoners against guards and other prisoners is a crucial goal. In many cases, the principal strategy for preventing such violence is the threat of the use of force by the guards. We have shown that it is diffcult to delineate contractually the permissible circumstances for the use of such force. Moreover, hiring less educated guards and undertraining them–which private prisons have a strong incentive to do–can encourage the unwarranted use of force by the guards. As a result, our arguments suggest that maximum security prisons should not be privatized so long as limiting the use of force against prisoners is an important public objective.
Basically, the case for privatisation is that incentives to hit performance targets encourage innovation. The case against them is that there are plenty of things (like inappropriate violence) that you can’t measure, yet help to make the performance targets you can measure. So the important question isn’t ‘what are the performance measures’, but ‘what can’t you measure’? If you can’t measure important aspects of performance and you give strong incentives to meet targets then there are likely to be unintended, and potentially unsavoury, consequences. That’s not an argument against privatisation but simply a caution that high-powered performance targets should be used with great care. It is particularly salient when we are talking about a contract that affects the physical and mental welfare of so many people.