Do higher wages reduce supervision costs?

Freakonomics points to the example of Hungarian musician entreprenuer Gabor Varszegi who made millions by providing high-quality service in his photo developing shops in Budapest by hiring workers and paying them wages that were four times the going rate. It quotes Hal Varian on the reason behind this apparently strange wage payments,

There are very few employees per store and monitoring their behavior is very costly.  If there were only a small penalty to being fired, there would be great temptation to slack off. By paying the workers much more than they could get elsewhere, Varszegi makes it very costly for them to be fired — and reduces his monitoring costs significantly.

In other words, higher wages act as a deterrent against shirking and working sub-optimally. Consequently, the need for more rigorous monitoring is less and productivity and retention increases. However, the success of this strategy lies in the threat of being immediately fired if they shirk. Remove this threat and the model collapses.

Public bureaucracies are the best example. Salaries at the lower levels of the permanent bureaucracy in India are very high, several times higher than the going market rate for similar work in the private sector. So, the Varszegi model should have made these workers more productive and reduced supervision costs. But a very strong countervailing force, in the form of (formal and informal) security of tenure or trade union activism, works against and virtually depletes the threat of being fired. The incentives get immediately misaligned.

So can this approach work with the contract employees in the bureaucracy. Thanks to minimum wages, they too get wages that are higher than the prevailing market rates. Theoretically, removing them is easy since the contracts can be easily terminated. And this is the case when the numbers involved are smaller. However, when such contract employees swell in number and have worked for a few years, the moral hazard resurfaces. They form unions and the lines of distinction with the regular employees, atleast to the extent of their labour rights, gets blurred. The threat of being fired gets diminished.  

In other words, the moral hazard arising from being part of the regular public bureaucracy or being part of large enough contract labour force who have been working for sometime nullifies the positive effect of the higher wages. As I blogged earlier, this is yet another example of how the emergent dynamics of human responses, under certain conditions, undermines logically sound incentive systems.