Rethinking and Rebooting

In 2013-14 , India has budgeted around Rs 200,000 crores ( roughly $ 30 billion) for its flagship programs - a significant amount intended to improve a range of basic services such as education, health, sanitation, child support, maternal health, rural roads, access to electricity, housing and guaranteed employment among others. But the results have been mixed at best and quite varied across states and districts. Worse , in some cases such as the mid day meal tragedy in Bihar where over 30 children died from food poisoning - the results have been tragic

The largest 10 programs in terms of outlay for 2013-14 are :

1. Mahatma Gandhi NREGA. Rs 33,000 crore
2. Sarva Shiksha Abhiyan Rs 27,258  crore
3. Pradhan Mantri Gram Sadak Yojana Rs 21,700 crore
4. National Health Mission Rs 21,239 crore
5. ICDS Rs 17,700 crore
6. Indira Awas Yojana Rs 15,184 crore
7. JNN Urban Renewal Mission Rs 14,873 crore
8. Mid Day Meal Program   Rs 13,215 crore
9. National Rural Drinking Water Program Rs 11,000 crore
10. RG Grameen Vidyutikaran Yojana Rs 4,500 crore

These programmes range from the Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) with an outlay of Rs. 33,000 crores to the Sarva Shiksha Abhiyan with an allocation of Rs. 27,258 crores. Food and nutrition schemes for children like the ICDS and the mid-day meal programme have been allocated over Rs. 30,000 crores. See the box below of the full list of the top ten programmes.

Various evaluations and assessments that have been done to improve the effectiveness of these programs show that some states such as Tamil Nadu, Gujarat, Karnataka, Kerala, Andhra Pradesh perform well on almost all the programs whereas some states such as Bihar, UP,Jharkhand, Madhya Pradesh and Assam perform much worse. With this emerging pattern of variability among states' performance we must step back and ask more fundamental questions. Why should we expect consistent performance on nationally designed programs with standard implementation templates across states with very big differences in implementation capability? Why should there not be different outcomes in a country with such huge diversity in state administrative and institutional capacity, in governance and in geographical and cultural differences?

This is especially true in the social sectors where the service being delivered has a high number of transactions with a lot of discretion. Providing health care or education would fall into this category of services. Clearly delivering quality education requires the coming together of a number of factors -- classrooms ( with toilets), qualified teachers, teaching material , appropriate curriculum, transport and mid day meals. Problems in any of these inputs would compromise the quality of education. So a lot of coordinated inputs, and de-politicized decision making and monitoring is needed to ensure that quality education is the outcome. Kerala has a long history and culture of providing good quality education to its citizens including and especially to its girls. Tamil Nadu, Karnataka and Maharashtra too have a good record in education. Gujarat still has some catching up – especially for girls education.

Similarly on health delivery trained nurses, medical materials and medicines and equipment, health awareness, accessibility to health centres, knowledge of basic hygiene and cleanliness, access to water and electricity are vital for reasonably good health outcomes. Coordinated and depoliticized decision making on a number of issues is needed to ensure proper health service delivery. If health workers (or anganwadi service providers) are hired with political connections to provide jobs to party workers then even if everything else is done right the system will fail.

In addition, social services like health and education not only have high transactions and coordination needs but also have low specificity - namely it's not easy to immediately measure their output. It takes time to measure whether a health or an education system is delivering quality health care or education and its hard to trace back what is missing if its not.

On the other hand building rural roads or electrification has somewhat lower transactions and can be contracted out through a bidding process if needed. It is also easier to measure and monitor output. Precise specifications can be required for rural roads or electricity lines, and monitors can come in quickly after a road is built to check if it meets specification, even before final payments have been made. The same cannot be done for education or health , other than by sending observers into classrooms and even then the learning outcomes may not be known for some time to come.

We need to think more smartly about how we will deliver education, health, or mid-day meals in a way quite different from how we deliver roads, electricity or even water. National schemes with very standardized specifications with very little flexibility in the way they are implemented will lead to very different results. It is therefore not surprising that the Pradhan Mantri Gram Sadak Yojana has a better record of implementation across the country than any of the social schemes.

The mid- day meal scheme is an interesting mixture of very high transactions and lack of specificity. It requires the meal to be cooked fresh on site, at a cheap price but not by a commercial company . At a minimum , it requires considerable local capacity to ensure that it can be delivered consistently and safely across thousands of schools. What is surprising is that we have not seen more tragedies of the type we saw in Bihar recently. Where civil society is strong and where parent associations are involved the scheme does well, but in remote areas and where these conditions do not exist a lot of corruption and misuse is the outcome.

So where do we go from here. One solution that has been suggested is to do away with national schemes altogether but transfer the funds to the states or to the Sarpanch at the local government level to leave it to their hands to provide the services. But by itself this will not solve the issue of variability of performance. The same issues which bedevil national schemes - weak institutional capability, local political capture, lack of accountability -- will devolve to the local level.

Instead we need to think more creatively While designing national programs which would have a more imaginative combination of standardization, flexibility and incentives. This would bring back the focus on better outcomes rather than a mindless obsession with a rigid set of national procedures that do not allow any innovation and also do not deliver services to people in many parts of the country.

Where its not easy to provide a cooked mid day meal safely to children might it be better to consider other options , such as pre-cooked, nutritious meals to children. Where anganwadi workers are being hired on a political rather than technical basis we must ask whether a different model for delivering health services to children and their mothers is needed. We don’t have all the answers but must allow more experimentation and piloting for different solutions to emerge working also with the help of good civil society organisations.

Local governance capacity is also critical for implementing the national programs. There appears to be considerable variability even within states. In Bihar which does not have a good overall track record in the past in delivering national programs Nalanda district outperforms other districts. Kullu district in HP is one of India's best performing districts on program implementation not only because of local administration but also because the elected Zila Parishads play a very proactive role in anticipating funds from national flagship programs and planning for their use with community involvement.

In states which have much weaker implementation capacity and basic institutional strength is weaker we need to build capacity in the programs as well as much stronger monitoring. In China the twinning approach is being used with more advanced states twinned with more backward ones. So its not just a centre – state issue but one of learning from each other. Might that work in India despite language differences ?

In any case we do need a more enlightened debate between the centre and the states and local government on how we change the focus from the funds for schemes and their monitoring to achievement of outcomes. We may also want to further rationalize the number of schemes to avoid duplication and overlapping mandates. The current top down model of national schemes being rolled out with pre-specified and rigid requirements and procedures is clearly in need of a rethink and a major reboot if India is to achieve its own national goals and the internationally accepted MDGs.