#346 Rebuilding

Cockroach Problems, India's Chip Future, and a New Tool for Policy Reasoning

India Policy Watch #1: Cockroach Problem; NEET Solution

Insights on current policy issues in India

—RSJ

Over the last month, we have once again found ourselves debating paper leaks, examination credibility, coaching centres, inflated educational claims and the ability of public institutions to administer high-stakes tests fairly.

None of these controversies is particularly new. What is disappointing is that despite their frequency and scale, they continue to be treated as isolated episodes rather than manifestations of a deeper problem. Every scandal produces a familiar cycle of outrage, clumsy investigations, arrests of a few coaching centre staff, the formation of committees, and eventual amnesia. The details change from one year to the next, but the pattern remains remarkably stable.

People seem to have already forgotten that we were in the same situation, at the same time, back in 2024. What came of that investigation? No one knows. So the pattern repeats. A leak occurs, or a website collapses, students suffer, explanations are offered, the troll army is pressed into service to hunt down detractors, and the system moves on. This time, we seem to have added another layer of pretence. The NEET test papers for the re-examination will now be transferred by the Indian Air Force. And the CBSE website will be evaluated by a team of experts from the IITs. The last few credible institutions in the country are now pressed into service to put lipstick on a pig. The obvious errors, like lack of planning and almost zero accountability, that stare at us from these episodes are barely addressed.

The recent CBSE evaluation controversy is a useful example. The transition towards on-screen marking was presented as a modernisation exercise. In principle, this was the right move. It makes the evaluation of test sheets location-independent for teachers, provides an audit trail for re-evaluation, and reduces logistical complexity. Large examination systems eventually have to digitise.

The problem was that the transition appears to have been executed without adequate preparation. Teachers complained about insufficient training and fatigue from doing on-screen marking. The scanning of such volumes of answer sheets was done without a proper test and control environment, resulting in mismatched answer sheets, lost pages, and often missed scans. Students reported discrepancies in evaluation after having a look at their scanned answer sheets. Re-evaluation requests surged, and the CBSE site struggled to stay up under the resulting demand.

The public discussion that followed focused largely on operational issues, but operational failures at this scale are rarely operational alone. How can such a large change be effected without pilots, a transition period and rigorous testing of each element of the chain? The number of discrepancies that have surfaced doesn’t suggest even a basic quality check before going live. A system that has not thought carefully about transition discovers them in production. That it involves millions of young students and their future should mean accountability for these lapses, and strong consequences should naturally follow. That is not visible at the moment. The government doesn’t fear public disaffection because it wins elections metronomically. The opposition is useless.

The same pattern is visible in the debate around NEET. Much of the discussion revolves around identifying the actors involved in paper leaks, determining the extent of the compromise and assigning responsibility for specific incidents. These questions are important, but they often obscure a more fundamental reality. The incentives involved in compromising NEET integrity are powerful. A government medical seat may cost a family a few lakh rupees, while a private medical education can cost well over a crore. The total number of seats in medical colleges in India is abysmally low for a country of its size. The NEET assessment model is still stuck in India of the 80s. A single examination, a single test paper of multiple-choice questions, determines the future of over 2 million students every year. The test paper is physical and static. The examination relies on a physical test booklet that must be printed, transported, stored and distributed through thousands of centres across the country. Under such conditions, leaks are not merely process failures. They are built into the design in which the rewards from compromise are enormous and the opportunities for compromise are plentiful.

This is why proposals that focus exclusively on punishment often feel incomplete. Disbanding institutions (such as NTA), arresting intermediaries, and strengthening penalties may satisfy the demand for visible accountability, but they do little to alter the underlying incentive structure. As long as the fate of a student depends overwhelmingly on a single examination conducted over a few hours, pressure will continue to build around that examination. The more useful reform proposals are those that attempt to distribute risk. Computer-based testing can be introduced gradually. Multiple testing windows can reduce the concentration of risk. Percentile-based normalisation can make large-scale administration easier. Most importantly, school performance can be given greater weight so that years of academic effort are not rendered secondary to a single high-pressure event. These ideas are all tried and tested modern testing frameworks. These might not eliminate all possible failures. But when a failure happens, these steps will ensure that the consequences are limited.

Yet even this conversation may be addressing only part of the problem. The recurring focus on plugging the leaks assumes that the examination is measuring the right thing and that the challenge is simply to preserve its integrity. Going through evidence collected by public-minded experts (Careers360 being one of them), the issue may be more complicated. Data from Maharashtra (which publishes such data digitally) revealed an extraordinary divergence between board examination results and entrance examination performance, despite both systems ostensibly evaluating the same underlying subjects. Students who excelled in one often performed unexpectedly poorly in the other. Such divergence does not automatically invalidate either system, but it raises an obvious question about what exactly is being measured. If two examinations covering the same syllabus produce dramatically different rankings, then at least one of them is not testing the right metric.

The coaching industry has understood this reality for years. They have turned into supplementary educational institutions with a particular specialisation. They have designed themselves to optimise for the entrance examination format. They teach pattern recognition, time management, elimination techniques, question prediction and psychological adaptation to high-pressure testing environments. When a student’s future depends disproportionately on a specific type of assessment, an industry inevitably emerges to optimise performance on that assessment. The remarkable growth of coaching should therefore be viewed not merely as an educational phenomenon but also as evidence that the examination system has become an end in itself. And over time, this industry has gone down the slippery slope of inflated claims, payments to toppers to retrospectively attach themselves to the institutes and eventually bribes and kickbacks to engineer leaks.

The comparison with China is also revealing, although perhaps not in the way it is often presented. China’s Gaokao examination involves more than thirteen million students, many times the scale of India’s major entrance examinations. It remains largely pen-and-paper, includes essays and long-form responses and is treated as a national undertaking requiring coordination across multiple levels of administration. The lesson here is not that India should replicate the Chinese model. The thing to learn is that examination integrity is fundamentally a governance challenge rather than a technological one. India often approaches these failures by searching for technological solutions, whether through digitisation, online testing or software platforms. Technology can certainly help, but technology cannot substitute for institutional capability. The harder questions concern planning, execution, accountability and the ability to learn from failure.

There is a broader issue that extends beyond examinations. The Indian education sector is one of the largest destinations for household savings and one of the most consequential determinants of life outcomes. Families routinely commit sums ranging from several lakh rupees to well over a crore in pursuit of educational opportunity. Yet the information environment surrounding these decisions remains weak. Coaching institutes advertise spectacular success rates. Colleges publish placement statistics that are often difficult to verify. Institutions compete through claims whose underlying methodologies are rarely transparent. And so you get Galgotias and many others who can get away with murder. Everyone within the ecosystem understands that selective disclosure, aggressive marketing and statistical embellishment are widespread. Yet regulatory scrutiny remains limited.

The state has the capacity to govern systematically important sectors and protect consumer interests. The contrast with financial services is striking. Over several decades, India has built a sophisticated architecture to protect financial investors. A person investing a relatively modest amount in a financial product benefits from disclosure standards, regulatory oversight, suitability requirements and enforcement mechanisms designed to ensure that claims can be independently verified. The burden of proof lies largely with the institution making the claim. In higher education, the situation is often reversed. A family committing twenty lakh, fifty lakh or even a crore rupees frequently has to navigate an information environment where verification is difficult and where the burden of challenging misleading claims falls largely upon the student or parent. It is difficult to think of another sector in which investments of such magnitude are made with such limited protections.

This asymmetry is not accidental. The entrenched political interests in the education system are largely responsible for this. We have devoted considerable intellectual and regulatory effort to protecting investors, while paying much less attention to protecting students and families making life-defining educational decisions. The result is a sector in which information quality is often poor, accountability is diffused, and incentives are misaligned.

It is against this backdrop that we saw the other phenomenon of this month, the emergence of the Cockroach Janata Party. The significance of the episode lies less in the satire itself than in the conditions that allowed it to resonate. Political satire succeeds when it captures a sentiment that already exists. The audience that embraced the joke was not responding only to an unfortunate remark made by the Chief Justice on NEET leaks. It was responding to a wider experience of institutions that frequently appear more comfortable managing narratives than confronting underlying weaknesses. Of course, it is a satirical website that will fade away soon and you could argue a whole lot of their followers are perhaps bots. Yet, it does capture something about the zeitgeist among young Indians who inhabit an environment shaped by examination controversies, unreliable information, recurring administrative failures and poor job prospects. Under such circumstances, satire becomes a vehicle through which accumulated frustration finds expression.

The speed with which the state responded to the Cockroach Janata Party by suspending its social media handles, calling them anti-national and sending out threats to its founders should be contrasted with the response on paper leaks and CBSE website failure. The state appears capable of mobilising considerable energy when confronted with criticism, symbolic opposition or perceived challenges to legitimacy. The response to structural failures that affect millions of citizens often is slower, more fragmented and less decisive. This is not a uniquely Indian phenomenon. Governments everywhere are tempted to focus on the visible manifestations of dissatisfaction rather than the conditions producing that dissatisfaction. People can accommodate mistakes if they believe institutions are learning, improving and holding themselves accountable. Confidence erodes when repeated failures appear to generate fewer consequences than those who draw attention to the failures. This is the India we are in now.

The unintended consequence is visible in the gradual widening of the gap between official assurances and public belief. Every unresolved examination controversy, every misleading educational claim and every unexplained anomaly contributes to that gap. Over time, trust becomes harder to sustain because citizens increasingly rely on lived experience rather than institutional communication. The real question facing us is therefore not whether a satirical movement is offensive, political or irresponsible. We have to ask instead if our institutions can rebuild credibility by demonstrating the qualities that ultimately matter most: competence, transparency, accountability and a willingness to learn from failure.

Without those attributes, the next joke is always on us. And, never very far away.

---

India Policy Watch #2: A Review of the Semiconductor Strategy Report by NITI Aayog

Insights on current policy issues in India

—Pranay Kotasthane

The NITI Frontier Tech Hub’s Future of India’s Semiconductor Industry report, released last week, is worth reading for those interested in India’s advanced manufacturing ambitions. It is a comprehensive take on this subject and gives some indication about the government’s future course of action in this domain. Hence, it’s worth taking a closer look at it.

What the Report Gets Right

There’s a lot to like about the vision the report proposes. As we have been writing in these pages, the core argument is that India should double down on its comparative advantages rather than just play catch-up. Here are the key paragraphs:

Winning the semiconductor race will not be easy if India continues to run the existing race; instead, it should shift gears and target becoming the ecosystem player that the global semiconductor industry cannot run without. This requires India to pivot away from the catch-up game in the foundry race and focus on winning the “More-than-Moore” era, where advanced packaging, system integration and manufacturing scale matter as much as transistor nodes.

By 2035, India should target building a USD 120-150 billion semiconductor value chain by choosing leadership and purpose over participation. Instead of chasing the global wafer race from behind, India should define its own pathway—one that is not only distinct but shaped by strategic self-sufficiency, ecosystem strength and global indispensability.

At the core of this vision is a resilient and disciplined manufacturing foundation, anchored by world-class fabs that focus on what matters most to India’s economy and strategic autonomy: mature-node logic, specialty analog and mixed-signal chips and compound semiconductors such as Silicon Carbide (SiC) and Gallium Nitride (GaN). Together, they will power India’s automotive, energy, industrial, telecom and strategic sectors—ensuring that technologies critical to India’s future are built on Indian soil. When it comes to advanced-node capability, the strategy most suitable for India is a selective and pragmatic approach aligned with national interest.

At the same time, India will play to its greatest strengths, including its design talent, highquality workforce and materials and chemistry ecosystem potential. Building on these advantages, it should aim to emerge as a global leader in semiconductor design and system architecture, a top-three destination for outsourced semiconductor assembly and test (OSAT) and advanced packaging and a trusted supplier of critical semiconductor materials, particularly wide-bandgap and advanced packaging materials. In these domains, India should aspire to set standards, shape supply chains and create enduring global dependence, rather than limit itself to mere participation. [Future of India’s Semiconductor Industry, NITI Aayog]

The report then goes on to identify five pillars for achieving these goals, all of which make quite a bit of sense. But I want to spend some time focusing on what I think are three gaps in this otherwise excellent report.

Gap #1: The Taiwan Contingency is Conspicuous by its Absence

The report organises India’s semiconductor imperative based on four reasons: import dependence, forex drain, national security, and societal upliftment. While these are legitimate concerns, the document largely sidesteps the most destabilising possibility in the global semiconductor landscape, a kinetic or economic disruption in the Taiwan Strait. This is not a tail risk, and is not limited to the most-advanced chips, as I have discussed in edition #324.

A report on deploying billions over the next decade should perhaps have an explicit section on what a Taiwan disruption means for India’s strategic and economic position, and how India’s investments can be structured to hedge against this scenario or position India as an alternative node.

While the report does discuss regular Multi-Project Wafer (MPW) shuttle runs, which aggregate designs from multiple companies and lower the cost of silicon prototyping, the destinations of these MPWs cannot be limited to TSMC alone. Encouraging these shuttle runs to happen at foundries outside Taiwan will provide a cushion to Indian firms if and when the Taiwan contingency happens. Even that won’t be enough. We will need a flexible sovereign fund to help domestic firms and start-ups absorb the two- to three-year drag they will face in the event of a Taiwan crisis.

Gap #2: The Design Section Misses a Trick

The report rightly identifies chip design as India’s strength. The Pioneering pillar includes sovereign EDA and MPW access, a National Design and Packaging Co-Design Platform, an AI-Enabled Semiconductor Engineering Mission, a National Frontier Semiconductor Research Programme, and a long-term Strategic Semiconductor IP and Patent Programme with explicit national IP targets and a sovereign IP fund.

But the report does not address the more difficult problems that have hindered the translation of Indian design talent into Indian intellectual property. The Design Linked Incentive scheme, announced in 2021, was structured around the same instincts—support for fabless companies, EDA tool subsidies, and IP filing assistance.

But overly restrictive Indian ownership conditions disqualified many firms from applying, and an unfavourable trade policy made importing second-hand design tools needlessly expensive. Moreover, difficulties in moving capital in and out of the country and the deterrent effect of tax policies mean that most Indian founders remain incentivised to create firms outside India. The key point being that pre-revenue deep-tech firms in India still don’t have access to capital markets.

The roadmap’s design recommendations make no reference to these bottlenecks. These issues are binding constraints for India’s success in most deep tech domains, not just in semiconductors.

Gap #3: Open-source Hardware is Absent

The EDA market is a tight oligopoly with Cadence, Synopsys, and Siemens EDA controlling roughly 97 per cent of the market. Tool licences run to hundreds of thousands of dollars per seat. This is a structural barrier to entry for Indian fabless startups. Besides, there is a geopolitical risk involved, given that these tools are controlled by a select few countries. Moreover, fabs prefer to work closely with a few of these players, locking others out of the game.

That’s why we made three recommendations in this Open Tech Strategy for India report:

• Co-create robust open-source hardware products with like-minded partners: Open-source Hardware need not always be Indian. Collaborations with like-minded partners will more effectively address the strategic imperatives for open-source hardware. Outputs generated by global open-source projects such as RISC-V will benefit everyone, including India.

• Award grants for creating the next generation of open-source EDA tools. Identify crucial toolchains required for chips used in critical applications and fund research to develop open-source alternatives.

• Make Semiconductor Lab (SCL) Chandigarh’s PDK for CMOS 180nm open access. SCL should enable Indian companies to fabricate prototypes of their designs cost-effectively. DARPA funded the Metal Oxide Silicon Implementation Service (MOSIS), which allowed US fabless start-ups to produce chips quickly. India, too, must enlist government research fabs in this project through a MOSIS equivalent fabrication service.

The NITI roadmap mentions EDA subsidies and AI-for-EDA pilots, and even proposes a sovereign IP fund to license ARM cores. But it says nothing about India’s participation in open PDK initiatives, nothing about OpenROAD, nothing about RISC-V as a design foundation for domestic startups. This is a missed opportunity because open-source alternatives are crucial for India’s strategic autonomy.

Overall, the NITI roadmap has many good ideas and a long list of recommendations for the government. But it takes the safe route of recommending ideas that come under the domain of MeitY. Converting India’s potential in this domain into geopolitical leverage, however, requires policy reforms across many ministries.

---

AIforPublicPolicy: A Policy Reasoning Assistant

—Pranay Kotasthane

For over a decade, The Takshashila Institution has built a public policy research methodology, i.e., a specific way of framing policy questions, using frameworks, encoding constitutional values, mapping stakeholders, tracing causal logic, etc.

Most of it lived inside the institution, in the heads of researchers, and in internal documents. We always believed it should be open access because every Indian state can have many Takshashilas. That was hard to do in the pre-AI age.

These are not proprietary secrets. They are learnable, teachable methods. We have now turned them into a tool called Scholar and released the entire stack as open source. The thinking behind making this open-source was that policy reasoning capacity is a positive-sum game. More people doing rigorous analysis makes the overall quality of public discourse better. If a researcher in a district collectorate, a journalist covering state finance, or a student at a university can use the same frameworks we use, that would be a terrific outcome.

The tool offers three entry paths. Start with a hypothesis and trace it through actor mapping and causal loops. Take a policy problem and run it through a structured framework to surface actionable alternatives. Or you can bring a completed draft and test it against Takshashila’s values. You can also use individual tools like budget analysis for Union and State governments, stakeholder map creation, and causal loop diagram generation separately.

Takshashila Scholar

Scholar is still in a beta version. Some tools are more polished than others. The workflow will change as we learn how people actually use it. Feedback is welcome and will shape what comes next.

Try it here. File issues or suggestions on the GitHub page.

---

HomeWork

Reading and listening recommendations on public policy matters

1. [Book] China’s Wars is a new book by former Foreign Secretary, Mr Vijay Gokhale. The book analyses five cases of warfare waged by communist China, and finds that in each of the instances, it was China that was the aggressor, even as it successfully crafted the narrative that it was only ‘counterattacking in self-defence.’ A must-read, and a quick one too.

2. [Puliyabaazi] Continuing with the China theme, this conversation ft. Manoj Kewalramani, Anushka Saxena, and Amit Kumar, discusses the future of India-China relations. Their book, Taming the Dragon, is a thought-provoking read. Don’t miss.

   क्या चीन के साथ दोस्ती मुमकिन है? India-China Relations: Cordial or Volatile?

3. [Report] This AI 2027 outlook is most fascinating and one of the most rigorous ones out there.

---