The first is: address delays in wage payments to restore the faith of workers in the programme. In 2016, the Supreme Court of India directed the government to ensure that wages were paid on time, calling the act of making workers wait for wages for months as equal to forced labour. However, there has not been even a single decisive step taken by the Ministry since then. Instead, the process of wage payments created by the central government has become even more convoluted. For instance, seven or more functionaries have to sign off before payment due to a worker can be approved (stage one of the wage payment cycle). This does not even include the series of delays from when the payment is approved till payment is made (stage two of the cycle). In contrast, the processing of loans from private banks is done in fewer steps. The point is the Ministry of Rural Development must simplify the payment process and has to be transparent about pending wage payments in stage one and two so that bottlenecks can be corrected. The second is: strengthen implementation capacities where expenditure is low instead of curbing expenditure where employment generation is high. States which are spending more are implementing the programme better because they have better capacities (as several studies including the government's own Economic Survey concluded in 2016). For a universal, demand-based social security programme such as NREGA, reforms cannot be based on targetting better. There has to be a focus on exclusion and not inclusion errors. Instead of using expenditure and income poverty as the only markers, exclusion must be identified at the household level. There is enough evidence to show that NREGA is fairly well targeted, benefiting the poorest, especially Scheduled Caste (SC) and Scheduled Tribe (ST) families. However, there is scope for improvement. For instance, panchayats, blocks and districts where employment of SCs and ST families is lower than their proportion in the population must be identified. This would indicate pockets where the most marginalised are being nudged out of the programme. Similarly, panchayats where the average wage being paid is lower than the notified wage rate must be identified as well. This would indicate places where the implementing authorities need to be hauled up for failing to ensure work is completed which in turn deprives workers of their minimum wage. The online Management Information System of NREGA can flag areas where entitlements are violated instead of being used as a tool by bureaucrats to centralise and control things. The third is: run the programme like a demand-based law, and not a scheme. Intermittent and unpredictable fund releases by the central government are one of the fundamental reasons why State governments are unable to ensure the full potential of NREGA. As of today, ₹18,191 crore in liabilities is due to 24 States. Poor performing States, on account of inadequate funds, typically discourage and often deny demand for work.