These concepts of regulatory episteme and fragmentation aren’t just theoretical and actively play out in real institutions every day. In our next post, we’ll see how this unfolds in the context of India’s competition governance. India has a young competition regulator, the Competition Commission of India (CCI), and an appellate tribunal for competition cases, the National Company Law Appellate Tribunal (NCLAT). In principle, both bodies deal with the same competition issues, but observers have noted an epistemic divergence between them. In simpler terms, the CCI and the NCLAT have developed somewhat different ways of looking at competition problems.

For instance, the CCI often prides itself on an economics-driven approach involving looking at market data, consumer harm, and the effects of business conduct with a fine-tooth comb of modern competition theory. The NCLAT, being a judicial appellate body, might emphasize legal interpretation, procedural fairness, and the sufficiency of evidence on record. These differing lenses can lead to tensions: decisions of the CCI have sometimes been overturned or modified by

the NCLAT, not just over factual disagreements, but seemingly because the two institutions weighed the issues through different frameworks. From the outside, it can seem like the two are not always on the same page about what competitive harm looks like or how to balance innovation against regulation. This is a prime example of a fragmented regulatory episteme in action.

When key institutions in a policy domain don’t share a common understanding, it can create uncertainty and inefficiency. Businesses under scrutiny might play one institution against the other, or they simply face prolonged uncertainty as a case ping-pongs between differing viewpoints. Moreover, the opportunity for learning for the tribunal and regulator to enrich each other’s perspectives can be lost amid turf issues or institutional pride. In the next part of this blog post, we will delve into this CCI vs. NCLAT dynamic. We’ll examine how their divergent episteme might have influenced competition law outcomes in India, and what that tells us about the broader need for coherence and feedback in regulatory systems.

Often, the impacts of regulation can be diffuse and delayed, spread across many players and over time. Regulators might not hear about unintended consequences until they’ve piled up, or they might only see anecdotal evidence. Without direct market feedback, an agency can only guess whether, say, a price control actually protected consumers or if it backfired by causing shortages.

Feedback from other institutions is also limited. Regulators don’t always have channels for sincere peer review or coordination where they reflect on each other’s outcomes. Instead of a culture of mutual learning, different agencies often operate in silos. At best, they might exchange formal comments or meet in inter-agency committees that often gloss over deep differences. Rarely is there a systematic loop where one body’s successes or failures inform another’s strategy in real time.

Even internal feedback within a regulatory agency can be problematic. Bureaucracies are not always known for self-critique. Frontline staff implement rules and might sense something is or isn’t working, but that information can get lost on its way up the chain. The agency’s leadership may be more attuned to political signals like pressure from elected officials or media criticism than to ground-level data.

When feedback loops are missing, regulators tend to move from one task to the next without learning. They issue a new rule, and instead of checking back in a year to rigorously evaluate its impact, they are already busy writing the next rule. Over time, this learning deficit means agencies stick with approaches they’ve always used, not necessarily because they’re the best, but because there’s no clear signal to change. In the worst case, a regulator might mistake silence for success (“no one is complaining, so it must be working”), when in fact the problems are simply unseen or unreported.