Policy Watch: Governance and Development – August 2026

This issue is on the theme Governance and Development. We carry three articles here. The first article, which carries significant excerpts from a paper by Kevin Grier and Robin Grier, both from Texas Tech University, USA, was published in the Social Science Research Network on 12th August 2026. Titled “Promises, Promises: Governance and Growth in India under Modi and the BJP”, it makes the point that “Modi’s rule is, at heart, a tale of broken promises.” In the paper, they show using counterfactual-based causal evidence how India has fared on both these dimensions.

Modi campaigned on good governance (“minimum government, maximum governance”) and on economic and developmental competence (the “Gujarat Model”). They show that for governance indicators including equality before the law, freedom of religion, freedom of association, and multiple indicators of democratic health, India dramatically underperformed its synthetic counterfactuals. The Indian economy underperformed as well. By the end of the sample, income was about 10 percent below its counterfactual.

Moving from the nation-state as a unit of analysis, we move to the second article by Apurva Kumar of RGICS, which makes an assessment of the Government e-Marketplace (GeM). Launched in August 2016, GeM has just completed ten years. By August 2026, the cumulative gross merchandise value (GMV) of procurement crossed ₹20 lakh crore, and more than 3.78 crore orders had been processed. The article presents a consolidated evidence-based overview of the principal objectives and policy intentions underlying the GeM. The assessment focuses on what GeM was intended to achieve and what the available evidence indicates regarding progress against these objectives.

The article concludes that there were measurable efficiency gains—nearly 10% cost savings as per a World Bank study. Improvements were reported in the timeliness of procurement. Supplier participation widened substantially, while the target level of sourcing was achieved from MSMEs and women-owned enterprises, though not from enterprises owned by socially disadvantaged groups (SCs/STs). There were incidents of cartel formation by suppliers, and genuine competitive behaviour still required vigilance. Digital controls reduced some discretion but could not eliminate human circumvention. Payment delays to suppliers remained a major issue, although significant improvements were recorded. In summary, we can see the enormous challenges of overcoming legacy problems, which require process reforms far beyond digitalization.

Next, we carry an article by Yuvraj Kalia, a former RGICS policy researcher. It is a review of the current Karnataka government’s interventions (May 2023 onward) for three important population segments—youth, women, and urban workers. The priorities of the youth are employment and earnings; of the women, it is ease of living; and of the urban workers, it is housing and mobility. The article details the schemes launched to address these priorities, starting with the “five guarantees”, their current uptake, and budget figures up to the 2026–27 state budget. In each of the three cases, the author identifies structural gaps that have not been addressed by successive governments and describes interventions that can be made to address those gaps.

Finally, we carry an article by Ashok Bharti of NACDAOR on the proposed Census 2027, which includes questions on caste for the first time since 1931, although a Socio-Economic Census was attempted in 2011. Unfortunately, due to the open-ended question “what is your caste” rather than a pre-defined set of responses with “any other (please specify)” as an option, the 2011 Census resulted in over 46 lakh caste names being collected. As a result, the Socio-Economic Census 2011 could not fulfil its basic objective of determining the differences across different social categories. Unfortunately, the Government of India’s announcement of the Census 2027 questions repeats the same mistake, and there is every possibility that a 2011-like fiasco may be the result. The article discusses how two states—Bihar and Telangana—conducted state-wide caste surveys with no problems because they followed a system where respondents first selected their caste from a predefined list, and only if their caste name was not included was it recorded separately.

As a tailpiece, we carry a table summarising the 50 main indicators from the National Health Survey, Round V and Round VI, classified by us into five categories. The good news is that only seven of the 50 indicators have moved in the worse direction, while 43 have shown improvements between 2021 and 2025. This table is part of an ongoing data analysis exercise, and RGICS will be doing more work on this in the future.