Democracy - An alternative perspective

In my last blog on democracy, the stand taken was that the democratic system is quite feasible and in fact could actually end up improving social welfare. Here I would like to offer an alternative perspective based on agency theory.

Theory: Let me put the theory in perspective. The mode of analysis will be based on the principal-agent model from information economics.There exist two entities in this world with the principal being the person with the capital and a project that needs an expert to implement and the agent being the person with the skills to do the job. The Principal hires the agent. The agent has superior information (since he is a specialist) which can be useful for the principal in the smooth running of the project. In return for the services of the agent, the principal provides him with a monetary transfer.

Problem: As can be seen straight away, the fundamental problem in such a model is information asymmetry between the principal and the agent. The agent has superior information which can be used to extract higher rents from the Principal. Since the principal knows this, he uses his powers of monitoring over the activity of the agent to keep the agent in check and controls the transfer in such a way that gives him optimal returns.

This leads us to the fundamental problem of efficient contracting. In a world where there is no information asymmetry and both the agent and the principal are aware of the entire information set, we achieve the socially optimal outcome where allocation of resources is in equilibrium. However from a practical perspective this is never the case and hence we will see some loss in efficiency. Such losses of efficiency can be attributed as agency costs.

Application: Think of the entire set of voters as principals and the candidate as the agent. THe principals are hiring the agent to represent them and execute public projects (we can treat everything as a project for illustrative purposes here) for them. The first point to note is that the principal does not have direct control over the actions of the agent (i.e.) the voters have to wait atleast till the next election to punish their elected representative and cannot nullify any actions he has taken on a project. Hence this is an obvious source for agency costs. In my story here, the transfers are in the form of taxes which the voters pay. The tax money is allocated by the representative for projects which may or may not improve social welfare. Therefore the representative can engage in wealth destruction for their subjects.

Who bears the burden or the costs? Does the representative bear the cost? My argument in the previous blog was that the voters can punish the representative in the next election. But consider this scenario. What if the representative decides he is not going to contest in the next election? What if he decides to make the most of his office now never to return later? The incentive structure now gets into trouble. Keeping that aside for the moment, the ultimate burden is borne by the taxpayers themselves. The projects are theoretically for the benefit of the public at large. However if the representative engages in wasteful uses of the funds, he does not bear any personal liability, but rather it is the taxpayer who ends up losing.

A logical question here is are there other forms of organizations where agency costs arise and if yes are there potential solutions we can look at? Think of a corporation as a counter example. Investors in companies also face similar problems. For instance if you want to invest in google, basically you are the principal and the managers in google are the agents. Obviously the managers at google know more about the firm and its prospects than you will ever know. Since this information is known to you, you can protect yourself through the market mechanism by demanding a higher rate of return for the google stock. Hence the premium collected on the stock covers the additional risk on account of information asymmetry.

Can we use this line of reasoning for the state? Probably not but it is interesting to think why not. In our little example here, the rules governing the uses of money and also the imposition of taxation is governed by the same authority. Hence the institutional set up is such that the principal is virtually powerless. In the corporate case, the principal can control his transfers to effectively get the work done. But he cannot do the same here. He has to pay the taxes imposed upon him by the state. NOw given there has been wasteful expenditure, can he ask for compensation from the state? Maybe you can but the chances are slim that you will get a ruling in your favour because of several reasons one of which being, preferences for public goods by consumers are never monolithic. The basic disagreement between the principals provides disincentives for public response (exceptions being different ideologies such as labour unions in kerala, where it really does not matter what u think. if the union decides to strike then all members go on strike) being effective.

Now getting back in line, I wanted to briefly mention about the comment made by brainwaves which can be directly linked to this story. As noted by brainwaves, there are times when the representative can strategically manipulate the incentives of the principals by attacking a set of principals with monetary offers and other incentives. This is a classic case where agency costs pretty much kill the system.

Even assuming that such strategic manipulations never happen, what can the public do? One option in a democratic set up, as i mentioned earlier in this blog and my previous blog, is that the principals ncan choose to replace the existing representative with a new agent. However the institutional structure still remains the same. Hence all we are doing is subsituting the agent but not his incentives. The new agent has exactly the same incentives as the agent prior to him. So why should he do anything different?

Under these circumstances does a democractic system really improve efficiency??

(PS: Being an empiricist by choice, I just cant stay away from saying this. Was just looking at the timing of the posts by Mindframes on Democracy (6:58), my post of decmocracy (9:59) and this last post (10:59)..now what is the significance of 59 in this episode. An innocous sense of timing one would think...ROTFL)