Autonomy for the CBI: Desirable but non-trivial

Economic Times, 16 May 2013

Autonomy for the CBI is a big issue facing the government today. However, when power is to be wielded by unelected officials, this cannot be done without commensurate accountability mechanisms. Getting the checks and balances to work correctly is truly hard. This calls for sophisticated expertise in public administration and law, and not simplistic drafting of law over a weekend.

Autonomy or independence is a valuable feature of public administration in five areas:

  1. Monetary policy: Politicians should not be able to rig elections by cutting rates in the year before elections. Therefore, we should have an independent central bank.
  2. Tax administration: Politicians should not be able to use tax enforcement to harass their rivals. Hence, tax administration should be autonomous.
  3. Criminal investigation: Politicians should not be able to use the police or the CBI to harass their rivals. Hence, the enforcement apparatus of the State should not be under the thumb of politicians.
  4. Infrastructure regulation: Big investments in infrastructure will achieve greater legal certainty if private firms do not face uncertainty across elections. Government is motherly towards PSUs; the regulator ought to be neutral between public and private. Hence, infrastructure regulation should be autonomous from politicians.
  5. Financial regulation: Similar considerations apply with financial regulation.

Thus, it is valuable for government to `contract out' work to autonomous specialised agencies in tax administration, monetary policy, criminal investigation, financial regulation and infrastructure regulation. Yet, the notion of autonomy or independence sits uneasily in the fabric of liberal democracy.

The foundation of accountability in liberal democracy is elections. Government is judged every five years, and every MP or MLA knows that winning elections is not easy. This brings accountability to the actions of politicians. We may disagree with what voters want, and we may disagree with how politicians read what voters want, but there is no question that elections are a powerful source of accountability. For the bulk of government, transparency and elections are the wellspring of accountability.

How do we achieve accountability for unelected bureaucrats? Bureaucrats do not face accountability through elections. Power corrupts bureaucrats as much as it corrupts politicians. Bureaucrats that wield power without accountability will be lazy, corrupt, and evil. If power is merely shifted from an accountable politician to an unaccountable bureaucrat, it makes things worse.

To achieve autonomy, four big issues need to be addressed:

  1. There must be clarity on the task. An agency with multiple objectives can always evade responsibility. Failure on one objective can be explained away claiming that another objective was being pursued.
  2. Political functions must be removed from the platter of such agencies. E.g. an independent RBI must not make a political decision about whether priority sector lending should go to a certain interest group.
  3. Actually achieving independence requires elaborate thought and care about the mechanisms of independence.
  4. Most importantly, elaborate accountability mechanisms are required, so that an independent CBI does not become a dreaded and unchecked secret police, with the ability to destroy the lives of citizens without checks and balances.

Analysing these issues, and coming up with sound laws, is a truly difficult problem. It is hard to make government work, and it is harder to make autonomous agencies of government work. All over the world, it took decades of thinking and experience to find the recipe for autonomy and accountability. Quick and dirty drafting projects such as the Lok Pal Bill, equipped with low knowledge of public administration, legal philosophy and international experience, will yield bad laws.

For an example of the scale of effort required, the Financial Sector Legislative Reforms Commission, headed by Justice Srikrishna, worked for two years and involved 146 persons in drafting the `Indian Financial Code' of which 30 persons were full-time staff. This draft law embeds an independent central bank and financial regulators. It specifies mechanisms of independence, clarity of purpose, and a subtle system of checks and balances through which unelected officials will deliver results.

The Indian Financial Code has done the job for two areas: financial regulation and monetary policy. Sections from the Code can be readily ported from financial regulation to infrastructure regulation. Hence, out of the five areas for work, we are roughly halfway there. Fresh thinking, on scale comparable to FSLRC, is required for the remaining two problems -- tax administration and criminal investigation.

Independence is in the air. Autonomy is a good thing in five areas: tax administration, criminal investigation, financial and infrastructure regulation and monetary policy. But sloganeering and weekend drafting will yield bad laws. Cogitation and expertise on the scale of FSLRC is required to battle the four challenges: (a) sharply clarifying the purpose of an agency that has to have autonomy, (b) removing political functions from its plate, (c) defining mechanisms of independence in the law, and (d) achieving adequate accountability mechanisms. With the draft Indian Financial Code in hand, we are halfway there.

Back up to Ajay Shah's 2013 media page
Back up to Ajay Shah's home page