The Lampost Framework: Why India Struggles With the Implementation of (some) Reforms

The ecosystem for implementation of reforms in India is structurally setup to solve acute, visible problems, but not chronic issues that require long-term monitoring.

By Akshay Alladi (@akshayalladi)

In much of our public policy discourse, many Indians are dismissive of state capacity. Much of what is run or managed by the state is shoddy- shabby hospitals, poor schools, crumbling roads and only intermittent power.

However, on closer examination, there are some areas where the Indian state’s performance is not just adequate, but indeed quite spectacular. Conducting elections in a free and fair manner, the eradication of polio through one of the largest public health programs in the world etc. are remarkable achievements.

Consider the case of polio eradication: The campaign was started only in 1995, and the total coverage of the target population was 99.7%! The WHO has now declared India to be totally polio free. Just a decade ago, the universal vaccination coverage in a state like Bihar was only 30%

What explains this seeming paradox?

If you look at it, a pattern emerges of the sorts of reforms the Indian state implements well, and what it doesn’t. The state manages to get several children into school, but fares very poorly on learning outcomes. It has been very successful in the eradication of diseases such as polio, but does badly on delivering healthcare in general. With the Mangalyaan mission it managed to reach Mars at an incredibly low cost, but struggles in delivering high quality science education to a broad mass of people. And as noted by Nobel laureate Professor Amartya Sen, the Indian state has prevented any famine from occurring in modern India (unlike in China or much of the developing world), but has a very poor track record on malnutrition.

The acute and the chronic

The pattern to note is that the Indian state does relatively well in handling “acute” conditions- that is those that require a specific intervention, for a limited time period, and with a clear, visible goal- which can measured at relatively low cost. The Indian state however struggles with chronic conditions- those that require painstaking management over a longer period of time, and where success is not as readily visible, so considerable cost and effort is required to measure progress.

The reason in some ways is the nature of Indian democracy. In Amartya Sen’s landmark work ”Democracy as Freedom” he asserted ”No famine has ever taken place in the history of the world in a functioning democracy”, and the reason he adduced was that democratic institutions—regular free and fair elections, independent courts and legislatures, free press and vibrant civil society—are all effective mechanisms of upholding the basic rights of citizens and would prevent a famine by providing effective feedback and pressure on the Government to act.

But why do the same mechanisms then not work in solving problems of a more chronic nature?

The lampost framework

To explain why reforms are difficult to implement in India (as opposed to why they are difficult to formulate and pass) I propose a new model (called the “lampost” framework). This framework  builds off the key concepts of Allison/ Elmore’s models as well as a modified version of Kingdon’s window specific to implementation (see schematic below). To illustrate the framework I use the case of sanitation or open defecation (OD) as an example.


Several initiatives, such as the recent Swachh Bharat, and the earlier Nirmal Bharat and Total Sanitation program (TSP) have sought to eliminate open defecation, but have progressed only on toilet construction, but not on the Information, Education and Communication (IEC) to improve toilet usage. Even now an estimated 600 million Indians defecate in the open, and only 46% of the toilets built in Year 1 of Swachh Bharat are reported to be used.

Explanation based on the framework: Absence of toilets is measurable at low cost, and building toilets is a one time activity addressing an acute issue (shortage of toilets). Hence, both for the media and for the public at large, by bounded rationality there is far greater emphasis on toilet construction and voters are rationally ignorant about toilet usage.

Though the media does highlight non-usage of toilets, such information is anecdotal, just given the high costs of gathering large scale information on toilet usage (a chronic condition). Hence, from a “demand” standpoint  it is easier for agenda setting on toilet construction (which then gets into the window of policy implementation), rather than usage (which is left out of the window).

The “supply” analysis is as follows: As a rational response to the “demand” side, both politicians and the bureaucracy prioritise toilet construction as a visible, measurable win; this is also because the allocation to IEC is lower (in fact it has been reduced to 8% of total funds in Swachh Bharat from an already low 15% earlier).

Given resource constraints the Government also cannot get a new, specialized implementation workforce focused on IEC- e.g., out of 76,108 Swachhata Doots required, only 8890 were recruited, the Communication and Capacity Development Units (CCDUs) that were supposed to implement this did not have dedicated staff, and had multiple objectives (Source: Arghyam Trust).

Hence the ‘bureaucratic actor’ who has multiple objectives, but not the commensurate capacity, rationally deprioritises the part that is less funded, and less measured- i.e., IEC. As an example of this behavior, in Himachal Pradesh IEC was initially prioritised with very good results for toilet usage, but as central allocation (and measurement) became far higher for construction, the bureaucracy prioritised construction, reversing the gains on sanitation.

The top down design of the sanitation program, also gave the line level bureaucracy very little autonomy or say in the policy design (as shown by the Himachal example)- hence from an Organizational Development standpoint the motivation to implement is lowered.

IEC and on-going toilet usage also depends on the last mile of the state- most of whose members are drawn from the same society who share the same prejudices about sanitation and are hence imperfect agents of change in social behaviour.

Finally, the activities of on-going maintenance and monitoring require coordination between multiple agencies. For example to build and maintain running water in the toilets, local officials must cooperate across more than 10 departments to obtain the relevant information, inputs and clearances as well as work with citizens and panchayats. These departments all have different objectives and priorities, and hence implementation for on-going maintenance is much more challenging.

I call this the “lampost” framework after the droll story about the medieval philosopher Nasruddin Hodja; when Hodja lost his keys he famously looked for them only under the lampost even though he likely dropped them elsewhere, because as he reasoned- what is the use of looking for something in the dark where it cannot be seen anyway! Much of the decision making in the Indian policy making is governed by the same principle- which explains the focus on visible wins that will be noted by the media, and hence the people, as opposed to the intervention that is likelier to have impact but is harder to measure.

This framework explains why India is good at solving acute issues/ crises/ one-time goals such as preventing famine (as Amartya Sen showed) or eradicating polio, but bad at implementing policies to address chronic issues that require sustained implementation and monitoring such as sanitation, malnutrition etc.

Akshay works in the e-commerce industry, and was a management consultant serving clients in the financial services and Government spaces. He is also an alumnus of the Takshashila GCPP13 Cohort.

, , , , , , , , ,

No comments yet.

Leave a Reply