Retreat in financial economic policy
At the dawn of the 1990s the founding idea in financial reforms was the promise of an exchange: You must give me money, I must give you securities. The perfection of the machinery, the certainty of settlement, this is what financial reforms was all about. We got to a world with day after day of flawless, sustained, settlement perfection. Writing in The Print, Bhargavi Zaveri and I worry about the loss of institutional memory and purpose.
Market failures and the case for SOEs
In “A Market Failure Framework for Evaluating Public Sector Undertakings”, published on The Leap Blog, Arjun Krishnan formulates a bipartite framework to evaluate state-owned enterprises, replacing the standard metric of profitability. This work is necessary because assessing Public Sector Undertakings (PSUs) via private-sector financial returns ignores their foundational mandate and confounds state decisions on privatisation or liquidation. Krishnan demonstrates that a PSU is ex-ante justified only when it addresses a persistent market failure that cannot be resolved through less costly interventions such as regulation, subsidies, or public-private contracting. The analysis establishes that ex-post performance must then be measured strictly through resource efficiency and the precise effectiveness of the enterprise in correcting its target market failure.
A cool district-level panel dataset
Two major developments in economic measurement are satellite imagery for night-time lights and for building volume. In each of these areas, there is a journey of establishing statistical and computational methods and then of innovating with creative applications. Examples of the former are link link and examples of the latter are link link.
Traditionally, the end-user required a significant amount of knowledge in order to get such databases up and running. For example, we have had many people like our bias-correction for nighttime lights data, and request this bias-corrected district level data that’s comparable with the World Bank release so that they don’t have to undertake this step. Similarly, many people are not yet using the superior Google building volume dataset.
In order to help reduce this friction for researchers working with Indian district-level data, we have released the XKDR Forum Nighttime lights and BV Dataset (2026). We have also released the code that gets to these datasets as Google colab notebooks, so you can reproduce our results, adapt this work for your country, or modify it in ways that fit you best.
XKDR Newsletter #50
The XKDR Newsletter for May 2026 is out.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.