## Tuesday, July 17, 2018

### Building State capacity for regulation in India

by Shubho Roy, Ajay Shah, B. N. Srikrishna, Somasekhar Sundaresan.

When India embarked on market oriented reforms in 1991, there was a desire to break with central planning; with detailed government control of entry barriers, product design and processes within firms. This is not synonymous with deregulation: there are market failures in many industries that require addressing. This led to the establishment of regulators. While the Reserve Bank of India has existed since 1934, there was a wave of establishment of new regulators after SEBI was created in 1988.

Regulators were to have legislative powers, to write subordinate legislation which would embed intricate knowledge and change rapidly with the evolution of fast-paced private industries. They were intended to have executive power in licensing and investigations. They were designed to shield transactions (licensing, investigation) from political interference. They were expected to achieve greater State capacity when compared with departments of government improving upon processes such as human resource policies.

The early optimism about shifting from central control to specialised regulators has given way to concerns about the working of regulators. Regulators have also been plagued by poor State capacity.

Regulators have too often veered into controlling as opposed to regulating, with creation of entry barriers and micro-management through regulations. Firms and groups of firms actively seek to co-opt regulators in their business objectives, which has given a return to the political economy of central planning. Entry barriers have sprung up with irrational and arbitrary decisions in licensing. Enforcement of regulations, and substantive law making, is selective and weak. This has induced large costs upon the economy. There is arbitrary power to initiate investigations and punish, and a climate of fear where private persons are afraid to criticise regulators. India of 2018 is uncomfortably similar to the India of 1991.

It is hence important to review the Indian experience, diagnose the sources of failure of existing regulators, and envision how high performance regulators can be obtained. We have a new paper, which is forthcoming in Regulation in India: Design, Capacity, Performance, edited by Devesh Kapur and Madhav Khosla, Oxford: Hart Publishing, 2019 (forthcoming). This article summarises and introduces this new paper.

There are two blind alleys in the quest for State capacity in regulators. It is possible to focus on one episode of a mistake by a regulator, and undertake analysis and advocacy about solving this problem. While a great deal of knowledge can be produced through case studies of failure, it is important to go upstream, to ask questions of incentives and information that lie at the source of repeated regulatory failure.

The second non-solution is a focus on personalities. When institutions are weak, the character of the institution is determined by its staffing. There is, then, a clamour to recruit great men, and then give them all power to do as they like (i.e. extreme regulatory independence). It is important to look deeper, to build institutions that have impersonal capability. For any change to be more than skin deep, it cannot be an idea in the minds of certain individuals; it has to be about the formal structures of governance, accountability, and processes.

Conversely, individuals working in regulators sometimes take criticism personally. However, the failures of an organisation are primarily induced by the design of the organisation. The same individuals would deliver superior outcomes if placed in a better designed organisation.

The focus must be on the incentives of the individuals who man the regulator. If regulators are merely given arbitrary power, public choice theory and the Indian experience shows that this power will be used poorly. What is required is systems of accountability, and checks and balances, through which the individuals working inside regulators have incentives to do the right things. This requires seven elements.

Clarity of purpose. Accountability for an organisation requires clarity on its goals. Every regulatory organisation must have a compact and clearly understood objective. Sprawling mandates, and particularly conflicting mandates, yield poor performance.

Role and composition of the board. The board must be dominated by non-executive members, through which the board can play the role of the Principal vis-a-vis the management which is the Agent. The board must control the organisation design, including organisation diagram, internal process manuals and the budget process. The board must control the legislative function.

Legislative process. When Parliament places law-making power upon unelected officials in a regulator, this calls for commensurate checks and balances in the process of regulation-making. The regulation-making process must start from the board. The staff must document the problem that is sought to be solved, the proposed intervention, and conduct a cost-benefit analysis. This packet must be put out for public comment. After this, the staff must address the public comments on the draft and make appropriate modifications to the draft regulation, and bring the draft back to the board for a discussion. Finally, only the board should have the power to issue a regulation.

Executive process. Sound processes are required in licensing and investigations, which protect citizens from arbitrary power. The non-award of a license causes harm for the applicant, and should use processes similar to those employed when punishing a citizen.

Judicial process An administrative law department should contain administrative law officers, who play no role in legislative or executive functions. This would yield an element of separation of powers. A hearing must take place before an administrative law officer, where the prosecution leads an argument and the defendant is given an opportunity to argue her case. This should lead to the drafting of a reasoned order as a structured document which shows the state of law, the facts of this case, demonstrates that law has been violated, or explain why this conclusion cannot be reached, and uses proper reasoning to determine the penalty. Orders should be published. There should be the possibility of efficacious appeal at a court or tribunal against the order. These three stages of due process (internally, at the administrative law officer, and externally, through publishing orders and at the appeal) create pressure upon the investigation and prosecution staff of regulator to produce high quality work, and protect citizens from arbitrary power.

Reporting and accountability. Regulators should be obliged to release statistical details about their functioning. Reporting should not concern the broader economic environment, e.g. the fluctuations of the stock market index, but should focus upon the actual work of the regulator, e.g. the win rate at the tribunal when orders are appealed. High quality reporting of all aspects of the working of the regulator will create the pressure of accountability, and feed into the budget process where targets can be set and incremental resources allocated in a way that pursues those targets.

The role of the department. Alongside the creation of well structured regulators, there is a need to clarify the role of the department of government, and create capacity in the functions that the department has to discharge.

These seven elements need to be coded into the Parliamentary law. This can be done at the level of one sector (e.g. the draft Indian Financial Code that envisages a single good governance framework for all financial regulators) or for all regulators in the Union government, as was done in the US by the Federal Administrative Procedures Act in 1946.

When compared with these seven elements of the design of a regulator which foster high performance, the present Indian landscape contains large gaps. The regulators of today are defined by skimpy laws, which give arbitrary power to the management, and lack a Principal-Agent perspective upon the working of the regulator. The present legislative framework is grounded in the notion that regulators are good people and will work towards the welfare of the people. This creates poor incentives for good behaviour by regulatory officials.

The FSLRC, chaired by one of us (Justice Srikrishna) drafted the Indian Financial Code, from 2011 to 2015. This draft law embeds the key ideas of this paper. Across the Indian landscape, many experiments are now taking place in building State capacity in regulators. This paper provides a conceptual framework, and 140 sections in the draft Indian Financial Code provides the draft law, for this journey.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.