In a new edition of I4I conversations, Maitreesh Ghatak (London School of Economics) is joined by Karthik Muralidharan (University of California, San Diego) to discuss his new book, Accelerating India's Development: A State-Led Roadmap for Effective Governance. 

Also available as a podcast.

Karthik Muralidharan begins by sharing his primary motivation for writing this book: economists talk about what the government should do, but often do not ask if the government will implement the recommendations or if it even has the capacity to do so. In this context, the book presents a framing of incentives and constraints of the two key actors that constitute the State – politicians and bureaucrats. The basic problem is that returns to State capacity are typically realised over a long time horizon, and hence, the electoral cycle does not always incentivise such investments. At the same time, the bureaucracy is under pressure to deliver, even when they do not have the required resources. Yet, once elected, politicians do care about getting something right. Academics can contribute by shedding light on the missed opportunities that policymakers can focus on. 

The book claims that investments in State capacity can yield returns of up to ten times. In response, Maitreesh Ghatak questions whether vested interests may prevent win-win policy options from being executed – with every attempt to enforce a law or improve accountability, someone is expected to be worse off, and they will oppose the change. Yet, as described in the book, if an overwhelming number of people would be better off due to a policy, it creates political support to action it. The idea is to look at the equity-efficiency trade-offs and invest in areas where there is scope for improvement. 

The book presents a framework – which involves ideas, interests and institutions – for making sense of the policies that are ultimately adopted. The role of researchers is to populate the public space with ideas, so that these feed into the public discourse and the concerned political actors can respond to them. The advancement in knowledge over the past 20 years, in terms of micro datasets, computing power, and greater attention paid to causal inference, allows us to identify ineffective government spending. Considering the example of human development, most of the funds are spent on unconditional salary hikes for regular government workers, despite various studies demonstrating that this does not affect productivity. On the other hand, putting a second Anganwadi worker in the community has been found to have a dramatic positive impact on learning outcomes and nutrition. Therefore, even within the same budget envelope of a particular sector, funds can be shifted towards better purposes. 

From the perspective of economic systems, Ghatak lays out three underlying mechanisms to make the most of low-hanging fruits in policy: the first is technological change, an external factor that makes it possible to do certain things; the second is rigorous evidence on a better way of doing something; and the third is political competition. He shares his scepticism around randomised controlled trials (RCTs) – will the government adopt policies that have been proven effective, and will they work in the same way as in the experimental setting? 

Muralidharan explains that, rather than advocating for the scaling up of specific interventions, his approach to the book is to extract broader principles from the evidence emerging from RCTs. The emphasis, therefore, shifts on building resilient public systems and institutionalising ways of functioning to make the State more efficient. The book covers the need to build better data collection systems, personnel management, revenue and expenditure systems, and the interface between the State and the market. Although each of these areas has been explored in empirical research, the intent of the book is to bring together the effective organisational principles derived from these studies, which apply not just to the public sector, but also to the private and non-profit sectors. 

Ghatak points out that many of the incentives being tested in RCTs are not resistant to collusion: should the goal be to design incentives which, once implemented, are not easy to dismantle or bypass? This brings up a deeper issue of what kinds of institutional investments can be made to maintain the integrity of the system. Muralidharan suggests that the interventions that have the greatest staying power are micro-level reforms to improve the budget process. While inefficiency in government is often associated with corruption, he notes that this accounts for only a small share of the problem. Much of the inefficiency is rooted in institutional inertia within an understaffed bureaucracy. He gives the example of a staff rationalisation exercise as part of an annual budgeting process. While incentives matter, many line departments currently lack the motivation to pursue cost efficiency. He advocates for a nuanced approach to squeeze out inefficiency– pursuing improvements without doing away with useful functions. 

Ghatak brings up a concern raised in reviews of the book, which how State capacity can be enhanced without upsetting the checks and balances on the government. For instance, inefficiencies in income tax systems or Aadhaar may provide some buffer between the coercive powers of the State and vulnerable populations. However, Muralidharan argues that if we accept the democratic premise that governments should deliver on what voters want, then they need to have the capacity to do so effectively, and guardrails ought to be enforced to prevent mismanagement and exclusion. 

The conversation closes with a reflection on future research and thinking. While we cannot experimentally ‘randomise’ the roll-out of policies across Indian states, variations across states could be leveraged to observe how similar interventions play out in contexts with different governments, cultures, and institutions.

1 Comment:

By: Neha Prasad 23 May, 2025

One of the principal factors contributing to the suboptimal performance of Government-run social institutions is the chronic unavailability of robust data on key parameters such as performance indicators, financial flows, personnel deployment, and decisions. Contrary to the common assumption that Government institutions have easy access to internal data, while a significant portion of institutional time and human resources in position is consumed by the basic task of data collection and maintenance, owing to the absence of a streamlined data management system. While the availability and access to internal data depends on the full-fledged functioning of a data-keeping system and willpower of the Government in sharing the data. The second reason is often overhyped, and the first one is highly neglected.

Show more comments

Join the conversation

Comments will be held for moderation. Your contact information will not be made public.