The challenges with Telangana's farm income transfer experiment

The Telangana state government's decision to implement direct income transfer to all its farmers is surely a landmark in India's agriculture policy space. Its outcome will be very closely scrutinised over the coming years. 

In brief, the State government have decided to transfer Rs 4000 per acre per season for the Rabi and Kharif crop seasons to all the 5.83 millions farmers as part of the Farmers Investment Support Scheme (FISS) or Rythu Bandhu Scheme. The transfers are to all agricultural land owners, irrespective of whether the land is brought under cultivation. This amounts to an annual subsidy outflow of Rs 12000 Cr or 7% of the total government expenditure, and would cover from 10-30% of the cost of cultivation depending on the type of crop. Two good assessments here and here.

Arguably the biggest challenge in achieving the scheme's objective of reaching small and marginal farmers arises from its not covering tenant farmers, a category not legally recognised in the State. Such farmers may cover atleast a third of all farmers.

So we have a pioneering agriculture policy reform being unveiled, perhaps the single largest direct income transfer program to farmers anywhere in the world. As Neelkanth Mishra has very nicely argued, it is almost impossible to make any reliable assessment of the program. And evaluations commissioned by the state government will take years to provide any actionable insight, if at all, whereupon the die would have been cast, either in terms of success or failure of the reform. Other states too would have jumped the bandwagon and emulated Telangana, inclusive of all the program's failings. So what is the best that can be done to ensure that the reform is effectively implemented?

At the very outset, we need to acknowledge that one could not have done an RCT to have evaluated the program before its implementation. Foremost, we do not have the luxury of time and KCR (the Chief Minister of the State) would surely not have had the patience. Public policy reforms rarely ever, if at all, happen in a calculated manner affording the luxury of detailed planning. They invariably happen as mutations. In any case, a small pilot evaluation would not have been able to reveal any of the several general equilibrium effects possible - how much of the money used for consumption, how much for investment, tenant-landowner dynamics and the income sharing, impact on land values, incentive distortion and leaving land fallow, impact on food prices etc. In the circumstances, the best effort would have been a rigorous qualitative assessment. 

Once the policy option is exercised, then the challenge is to ensure its high fidelity execution. This would mean ensuring land ownership details are accurately captured (can remote sensing data and GIS mapping, coupled with a field-survey, help with a one-time clean-up?), payments processed and delivered to the farmers in the most cost-effective and most accessible manner (can technology solutions and digital money help?), some way (short of a regulation or rule) in which tenants can negotiate with landowners to get a share of this money (can nudges help?), the money is withdrawn and utilised in the most productive manner (again nudges, say, to purchase farm inputs?), the float in the distribution channel by way of locked up money due to deaths etc be minimised (can technology help?), discourage farmers who could leave land fallow and just collect the transfers (some information disclosures and structuring of the payments be of use?) etc. 

As can be seen, each of these problems can have unique ways to address them. The government would need to innovate improvise continuously. 

Addressing these execution challenges and ensuring that the reform realises its full value would require action at three levels.

The first would be purely at the level of execution management. Can we have a monitoring system with tight feedback loops that inform decision-makers at District and State levels about bottle-necks, distortions and problems as the implementation proceeds (say, larger and absentee farm owners leaving land fallow to collect the transfers which is higher than the tenancy rent)? Can there be a back-up team which can respond to such emergent concerns and address them swiftly, both at the policy level as well as, more likely, at the level of field implementation? Can we have a strong analytics team that is able to rigorously analyse the data exhaust and offer actionable insights, which can perhaps help iterate and improve the policy over time?

The second would be at the level of policy elements. Can the State government emulate the Giveitup campaign associated with the LPG subsidy program of the Government of India and nudge the richest farmers to voluntarily abstain from taking the subsidy? What would be the most effective way to exercise such moral suasion? Given that 9% of farmers with more than 5 acres each own a third of the land and therefore would claim a third of the subsidy, can the government go one step further and cap the subsidy in an administratively simple manner? Going forward, as the farmer database and transfers distribution channel stabilises, can the government explore options of targeting farmers, crops, regions etc? 

The final level of engagement would have to be at the eco-system. Gradually, after a year or so of the implementation, can the database and monitoring system be used to deliver other types of services? How does this work-flow integrate to the fertiliser subsidy transfer system? Can the foodgrain procurement process be linked up with this database? Can this be linked to the agriculture e-market place, eNAM? Can we use this digital spine to deliver direct cash transfer in return for erecting meters on agriculture power connections? Can we gradually build a robust agriculture information management system and a platform to deliver various kinds of farm services?

It is all too easy for me to write these down as a sort of pre-mortem. In fact, I could dig deeper and get more granular at each level. It is an altogether different task for the State government to just keep its eye on all three levels always, much less translate them into action. The best that can be expected is for the State government to perform reasonably well the limited task of high fidelity execution. That is a two-year agenda.

For now, despite all its flaws, the state government should be applauded for the leap of faith, as is the case with any such reform.

It does not need any great foresight to assess how the program will get implemented, if it is done business as usual by the State government. It is almost unrealistic to expect a state government, even a very high capacity one at that, to execute at all three levels. Even high fidelity execution will be a great achievement. The ebbs and flows of political cycles alone are enough to disrupt any neatly laid down plans at policy and eco-system levels. No point in criticising State government for such failings. We only need to be surprised if that does not happen. 

None of this would prevent opinion makers and academic researchers from sitting judgement five years hence, and with the benefit of hindsight smugly castigating the government for all the distortions and failings (some which cannot even be anticipated now) that would have inevitably crept into the implementation - it was not evidence-based policy making, there was corruption, there was no political commitment, the tenants and therefore the poorest farmers did not benefit, and so on. We told you so! 

Instead, the challenge is to engage in real time. Can evidence-based policy making ideologues offer the State government something tangible in terms of the aforementioned engagement elements that increases success likelihood as it embarks on this challenging reform path? Anyone up for that challenge?