Incentive system facing IAS officers

It has been a widely held view for a long time that the highest bureaucracy in India, represented by the Indian Administrative Service (IAS), has become captive to the political system, thereby adversely affecting the morale of the service and the efficiency of public service delivery. It is no secret that politicians use posts
of varying importance to motivate bureaucrats, who respond by investing in political loyalty or (in fewer cases) skill to get important posts. The most visible manifestation of this phenomenon is the now commonplace large numbers of transfers that takes place when governments change after elections.

However, a fascinating BREAD paper by Lakshmi Iyer and Anandi Mani, which explored the relationship between IAS officers and the political system, arrives at a few interesting (and heartening) conclusions:
1. High-skilled officers face less frequent transfers and lower variability in post importance due to political change.
2. There is no robust evidence of a negative impact of such politically-induced transfers on district-level policy implementation outcomes.

The analysis is from a simple principal-agent framework, where the politician's (principal) electoral success is a function of the bureaucrat implementing policies effectively and bureaucrats (agent) aspire to hold important posts. In the absence of direct control over the bureaucrats, the politicians respond by using transfers and postings as a leverage over their agents. The bureaucrats respond very early in their career, either by developing a reputation for expertise in a specific sector, or more commonly by pitching their loyalties to one or a group of politicians or political party.

The article validates the following hypotheses empirically, by tracking the careers of over 4000 IAS officers across India over the 1980-2004 period
1. If the electorate cares not just about efficiency, but also about redistribution issues, political change will result in reassignments (transfers) of bureaucrats across posts.
2. Junior bureaucrats with high initial ability are more likely to invest in developing expertise as their route to career success.
3. High-skilled bureaucrats are less likely to be reassigned to new posts as a result of political turnover than bureaucrats who develop a reputation for loyalty to particular politicians.
4. Political change will cause less variability in the importance of posts held by these high-skilled officers, relative to loyal bureaucrats.
5. Since politicians main motivation for such transfers is to influence their re-election prospects, political change is more likely to result in transfers among posts which a¤ect a greater number of voters.
6. Political change is less likely to result in bureaucratic turnover if alternative means of controlling bureaucratic outcomes are available.

The other empirically validated findings of the study, many of them very obvious, include the following
1. The study also finds that that those with high initial ability are more likely to be recommended for senior positions in the central government in New Delhi, suggesting the Central Government bureaucracy is more professionally competent than the State Government one.

2. The finding that officers who develop professional competence are no more likely to get better postings (over an entire career) than those pitching their tents behind political leaders or parties, is an incentive to officers to take the more easier route to "success" (high-profile postings) by investing in political loyalties.

3. Officers who belong to the same caste as the Chief Minister's party base are more likely to be appointed to important positions.

4. District Collector and other district level positions that directly impinge on the interests of the local politicians and parties, are more vulnerable to transfers when governments change. More populous districts are more likely to witness bureaucrat transfers following political change, consistent with the politician's motivation to use bureaucrat reassignments to inflence a larger number of voters.

5. Bureaucrats in districts where the local politicians belong to the same group (in the party) as the incoming Chief Minister are less likely to be reassigned by the latter. This is consistent with the model's prediction that the availability of alternative means of control over bureaucrats lowers the politician's incentive for transfers. In other words, this suggests that the Chief Minister regards local politicians and bureaucrats as substitutes for exerting their influence at the district level.

My strong contention is that the bivariate (loyalty and ability) analysis is simplified in considering the two variables as mutually exclusive. The game becomes fiendishly complex, once we have loyal officers also being capable and vice-versa, as is increasingly the case. This deserves more anlysis and will be the focus of more posts in coming days.