<b>Ajay Srivastava:</b> A rating system to transform governance

Policies with high ISR scores will radically improve the government's interface with the citizen, making it quick, transparent and free of corruption

Let us consider two variants of a policy. The first says VAT on mangoes would be payable at five per cent of the value sold. The second prescribes VAT at the rate of eight per cent in general, but five per cent on tropical fruits used for making squash, by a unit located in backward regions and not blacklisted under any Act. In the second case a person selling mangoes will need to submit many documents and follow a tortuous path to justify paying VAT at five per cent and not eight per cent. We can see that the second variant is more likely to fail, as it is difficult to implement.

When such policies fail, senior bureaucrats blame field officers for bad implementation, but as the above example illustrates, most of the time a policy fails because it contains faulty implementation genes. Is there a way to diagnose such genes in advance?

One way would be to introduce an "implementation success rating" (ISR) for public policies. ISR would be a number signifying how best a policy can be implemented. The higher the number, the better will be the implementation.

The ISR system can be used to rate all public policies and schemes on a scale of zero to 10. A policy that can be implemented through an automated system with the least human interface will get an ISR of 10, while a policy where implementation is a nightmare will be rated at zero. If a field officer, while processing an individual request, is required to apply his mind on translating policy intentions, it's a bad policy and its implementation will be characterised by delay, use of discretion, and corruption, and would require a large field staff.

Let us obtain the ISR for the two policies we referred to above for refund of VAT. The policy that allows users to pay VAT at five per cent online will score a perfect 10, while the policy that requires payment subject to numerous conditions would have an ISR of two or three. This is not to say that policies should not contain multiple conditions. All we say is they must incorporate as many conditions as required, but in such a way that they are amenable to programming and hence automated disposal. In the example of mangoes, if the policy says that VAT is levied at five per cent on mangoes, guavas and litchis only, instead of using the expression "tropical fruits", and also names backward areas and Acts, the ISR would jump to a higher range. Here the policy remains the same, but has been defined more precisely and because of this, it has become clearer and amenable to programming.

Rated on this criterion, the filing of income tax refunds for individuals will qualify for an ISR of 10, as it allows users to complete filing through a simple online interface. Calculation of import duty today would merit an ISR of three, as one may need to refer to numerous notifications which defy programming.

The rating can be implemented in four easy steps. First, government should announce the adoption of ISR as one of the major tools for assessment of public policy. Its most successful polices such as Jan Dhan, Aadhaar and LPG subsidy refund already have elements of star ISRs. Second, an agency may be entrusted to prepare detailed ISR guidelines. The success of ISR will depend on clear and undisputed definitions of the attributes applicable for each rating from zero to ten. Third, all departments must rate existing policies and publish policy-specific ISRs on their websites. Fourth, all departments must then move aggressively towards modifying these policies for achieving higher ISR scores. This would require user-centric application development and complete business process change in many cases. No new policy should be adopted for implementation unless it is certified as having a star ISR rating.

A new class of external auditors would be required for this task. These auditors will help government agencies in rating existing policies and then help departments to move towards higher ISR scores.

Some commonsense practices greatly help in automation. Draft a simple self-contained policy with minimal need to refer to too many conditions or external documents. Make all rules computer-interpretable. We have a large number of rules that cannot be automated because of ambiguity or complexity. Rules that cannot be made computer-interpretable must be amended. No new rule should be passed unless it is certified as computer-interpretable. This will remove scope for multiple interpretation, delay and discretion later. With the introduction of computer-interpretable policies, most decision making can be made automatic.

Why is improving a policy's ISR important? Policies with high ISR scores will radically improve the government-citizen interface, making it quick, transparent and corruption-free. ISR will also make senior bureaucrats accountable, allow them to take credit or be held responsible in a transparent manner. This will foster healthy competition in improving public services.

ISR will be a sure and inexpensive tool to promote maximum governance through minimum government.
The writer is from the Indian Trade Service. These views are personal