The Nitish Kumar government has proposed that the electorate be given the right to recall corporators. This is yet another of recent experiments from Bihar, after the cash-transfer schemes for bicycles and school uniforms, and the abolition of the MLA Local Area Development Scheme. We discuss the merits and conditions under which a recall system could be effective.

The Bihar proposal is a modification of an existing right to recall councillors. Under the current Bihar Municipal Act, a councillor can be removed on a petition signed by two-thirds of his fellow councillors. Under the proposed system, this power is given directly to the electorate — if two-thirds of the registered voters of a constituency sign a petition to the urban development department, the government can take steps for removal of the corporator.

Madhya Pradesh and Chhattisgarh have also permitted recall of corporators. The Chhattisgarh system differs from the Bihar proposal. The process can be initiated against a corporator who has completed two years in office. Three-fourths of fellow councillors have to write to the district collector asking for a recall election. On the recommendation of the district administration, the state election commission holds a “confidence” vote in the constituency. If the majority of the voters vote against the councillor, he is removed. In 2008, three presidents of municipal councils were removed through this procedure.

The main argument for the right to recall is as follows. The people of a constituency decide who represents them in the corporation, state assembly and Parliament. The representative has to be held accountable for their actions. During the term of the representative, there is no way of ensuring performance accountability. The electorate has that power only at the time of the next election. There could be merit in providing the electorate with the power to ensure accountability during the term.

The main argument against the power to recall is that this impedes medium- and long-term policy implementation. Many policy decisions have a gestation period for implementation gains, and could have short-term costs. Representative will take such decisions only if they are confident that they will not be penalised for doing so. Having the assurance of a four- or five-year term permits them with the leeway to take such risks. This is the reason most national governments take tough decisions in the first couple of years of their term — witness President Obama’s health bill and the British welfare cuts. As an analogy with the corporate world, managements focused on quarterly results may often underperform than those with a longer-term shareholder view; this distorted the incentive structure for derivative traders and exacerbated the systemic risks that led to the global financial meltdown in 2008.

The recall system needs several conditions to work. First, it should be recall for individual non-performance, and not because a political party has become unpopular. Otherwise, even a well-performing representative can be penalised for some decisions taken by his party. It may be difficult to ensure this distinction in cases where a representative is elected on a party ticket. Second, there should be a minimum grace period to judge the performance — Chhattisgarh does not permit recalls for the first two years. Third, the process should be cumbersome enough to deter casual usage. The Chhattisgarh law does this by requiring a three-fourth majority in the corporation to initiate the process. In Bihar, the proposed law required two-thirds of the electorate to sign a petition — a high barrier, given that most elections do not even see that level of turnout.

The issue is more complicated for MLAs and MPs. The anti-defection law mandates that they vote on party lines. There could be instances when the party mandate is contrary to local interests. This would place the MP (or MLA) in the difficult position of choosing between the risk of being disqualified for defection, or being recalled by the people of his constituency. This example highlights the tension between accountability to the electorate and to the party. The recall system could work better in small constituencies in which members are elected as independents (without explicit backing from a political party). For example, it could work in municipal corporations in Uttar Pradesh, as all corporators had to contest as independent candidates.

Democratic processes have to evolve, adjust and respond continuously to new challenges. Given our federal structure, states are an ideal place for experimentation within the overall constitutional scheme. The recall processes in Madhya Pradesh, Chhattisgarh and Bihar will hopefully throw up learnings that can be utilised by other states considering a recall system.