Too Strategic to Fail: The Political Economy of India's Defence Public Sector

The Puzzle: Failure without Correction

In 1989, India sanctioned the Kaveri jet engine program. Over the next three and a half decades, the state spent roughly ₹2,839 crore attempting to develop an indigenous fighter engine. The program produced laboratories, testing infrastructure, technical research, and engineering expertise. What it did not produce was an operational engine. No aircraft in Indian service flies with the Kaveri.

Despite this, the program was never officially cancelled. No leaders were publicly held accountable, and no serious review examined why decades of spending yielded so little usable capability.

The puzzle extends beyond the technical difficulty of jet engine development. Advanced defence projects fail even in technologically advanced states, and India faced genuine constraints including technology denial regimes and limited industrial depth. These factors help explain underperformance. They do not explain why underperformance produced so little institutional correction.

Democratic systems are generally expected to discipline persistently failing institutions through political pressure, bureaucratic competition, or customer dissatisfaction. India’s defence public sector has largely escaped all three. Hindustan Aeronautics Limited (HAL) and the Defence Research and Development Organisation (DRDO) remain central to India’s defence industrial structure despite repeated delays, production bottlenecks, and capability shortfalls extending far beyond Kaveri alone.

Therefore, the central question is why visible institutional failure has produced so little restructuring.

This article argues that India’s defence public sector persists because the costs of accountability are concentrated and immediate, while the costs of failure remain diffuse, delayed, and strategically manageable. Electoral geography protects DPSUs locally. Internal bureaucratic incentives reward research activity more reliably than operational delivery. Inter-institutional adaptation reduces the incentive for any major actor to force confrontation. Under such conditions, institutional persistence becomes politically safer than institutional correction.

Existing scholarship, particularly Anit Mukherjee’s Absent Dialogue, identifies the fragmentation and vested interests embedded within India’s defence management system. This article focuses instead on the mechanisms through which those dynamics survive across governments, procurement cycles, and strategic crises.

The Electoral Constituency Force Field

The persistence of India’s defence public sector cannot be understood through organizational analysis alone. DRDO and HAL are not bureaucratic entities operating outside electoral politics. They are employers, contractors, housing anchors, and middle-class ecosystems embedded within parliamentary constituencies.

Once viewed through that lens, institutional persistence appears less anomalous and more politically intelligible. Most analyses note where defence public sector facilities are located. Geography alone, however, explains little.

The more important distinction lies in electoral competitiveness. A Defence Public Sector Undertakings (DPSU) facility situated inside a politically secure constituency generates different incentives from one located in a closely contested seat where organised employment blocs can influence outcomes.

As such, the relevant question here is how their employment networks interact with constituency-level political risk.

HAL’s major production centres stretch across Bangalore, Nashik, Koraput, and Lucknow. DRDO laboratories cluster through Hyderabad, Pune, Chennai, Delhi, and Dehradun. Their political significance exceeds raw workforce numbers. Employees are often socially concentrated, unionized, civically active, and embedded within local supplier and service economies. In competitive constituencies, such networks can exert influence disproportionate to their numerical size.

Lucknow illustrates the mechanism clearly. HAL maintains a significant industrial presence and ecosystem in the Lok Sabha constituency represented by Defence Minister Rajnath Singh. The institutional tension is structural rather than personal. The minister, formally responsible for defence modernisation, simultaneously represents a constituency whose economic ecosystem is tied to HAL’s continued stability and funding flows. Aggressive restructuring would impose immediate and visible political costs within his own electorate.

The same incentives would confront almost any politician occupying a similar position. Electoral systems reward responsiveness to concentrated interests more reliably than promises of long-term efficiency gains whose benefits remain diffuse and uncertain. The structure matters more than the individual within it.

The political footprint of a DPSU extends well beyond payroll figures. HAL and DRDO facilities sustain subcontracting chains, retired employee communities, educational institutions, and local housing markets that have evolved around them over decades. In several constituencies, the defence establishment functions as a source of middle-class prestige and local identity as much as employment. Organized employees and affiliated networks also vote at higher rates and mobilize more effectively than the diffuse beneficiaries of reform.

This protection structure is reinforced by its distribution across party lines. DPSU facilities are spread through constituencies represented by the Bharatiya Janata Party (BJP), Congress, and regional parties alike. Over time, defence-industrial placement became politically embedded across the wider party system rather than tied to any single ideological coalition.

Electoral turnover therefore does little to weaken institutional protection because successive governments inherit similar constituency incentives regardless of political affiliation.

The logic resembles Mancur Olson’s argument regarding concentrated interests and diffuse costs. DPSU employees understand clearly what reform may threaten and possess the organisational capacity to resist it. The beneficiaries of reform are far less politically visible. A pilot waiting for improved capability, a taxpayer absorbing procurement inefficiency, or a future government inheriting capability gaps does not constitute an organised electoral bloc capable of imposing immediate political penalties.

Institutional protection thus emerges from the ordinary functioning of democratic incentives themselves. That is precisely what makes the system so durable.

The Big Misalignment

The persistence of DRDO’s performance gap cannot be reduced to electoral shelter or political caution alone.

Even if those constraints were loosened, the organization would still be governed by an internal reward system that treats measurable research activity as success, and operational deployment as a secondary or delayed concern. The misalignment is not rhetorical. It is built into how careers advance, how prestige accumulates, and how institutional memory is recorded.

DRDO’s internal evaluation structure privileges patents, publications, laboratory milestones, technology readiness progression, and prototype demonstrations. These are observable, reportable, and defensible within administrative review.

Operational deployment, by contrast, sits at the end of a long chain of approvals involving the services, production agencies, and procurement committees. It is therefore easier to treat as an external contingency than as the central measure of institutional performance.

Over time, this produces a quiet substitution. Intermediate outputs cease to function as steps toward deployment and begin to function as the object of work itself. The organization does not simply drift away from its stated mission. It begins to define the mission through the metrics it can reliably control.

The Kaveri engine program illustrates the structure with unusual clarity. Across three decades, it generated real technical progress: turbine design work, materials research, testing infrastructure, and an accumulated base of engineering expertise.

Within DRDO’s evaluative framework, these outcomes were not peripheral. They were evidence of continuity and advancement, sufficient to justify continued funding and institutional renewal.

Yet from the standpoint of the Indian Air Force, the central fact remained unchanged. No operational aircraft entered service powered by the engine. The divergence between these two assessments is not a matter of interpretation. It reflects two different systems of accountability operating on the same program without a shared closure condition.

The deeper institutional issue lies in how they are interconnected. DRDO does not operate under a contractual model in which the services can impose penalties, terminate programs unilaterally, or enforce binding delivery benchmarks. Instead, accountability flows through administrative and ministerial channels where research progress is more legible than operational absence.

A program that fails to deliver can still generate annual evidence of advancement. A capability gap, by contrast, often appears only at the moment of operational need.

Tenure fragmentation reinforces this structure. Program leadership rotates every few years while major development cycles extend across decades. Decisions are made by individuals who will not be present when outcomes fully materialize. As a result, responsibility is distributed across time rather than concentrated in decision points. Instead of continuity, this becomes a sequence of unlinked accountability windows.

A comparison with the Indian Space Research Organisation (ISRO) clarifies what is structurally distinct. Both institutions share a state-led scientific heritage, comparable talent pools, and similar funding origins.

Yet ISRO operates in an environment where performance outcomes are externally legible and difficult to reinterpret. A launch either succeeds or fails. The evaluative closure is immediate and public. DRDO programs operate in a more elastic space, where technical progress can be continuously demonstrated without requiring definitive operational transition.

This difference is not cultural but architectural. ISRO’s metrics are tied to irreversible outcomes. DRDO’s metrics remain internally negotiable, allowing programs to remain perpetually “in progress” without confronting a hard endpoint.

The result is an institution that is not failing to produce work but steadily producing the wrong kind of accountability signal. Over time, research activity becomes both the evidence of success and the justification for continuation, while deployment becomes an external problem rather than an internal obligation.

The Avoidance Equilibrium

The persistence of underperformance in India's defence public sector is best understood not as a result of isolated failure but as a stable pattern of adaptation across the armed services, the Ministry of Defence bureaucracy, and the defence public sector undertakings.

Each actor responds rationally to its immediate constraints. Taken together, these responses produce a system in which confrontation is consistently more costly than accommodation, and ambiguity becomes the default mode of governance.

DRDO and HAL, in turn, operate within an environment where programs are rarely cancelled and delays carry limited consequence. Institutional survival depends less on final delivery than on continuous demonstration of progress. Patents, prototypes, and revised timelines become proxies for success.

Over time, these intermediate outputs begin to substitute for operational deployment itself, a pattern documented by the Comptroller and Auditor General, which found that projects were closed as "successful" despite the non-achievement of key objectives or parameters, redefining achievement in terms that the system can reliably measure and defend.

The Ministry of Defence occupies the central mediating role. It must balance the services, scientific establishments, and political sensitivities surrounding public-sector employment. Direct confrontation is costly, while procedural management through committees and extended reviews is safer. Cancellation or termination creates visible political exposure, so disagreement is absorbed rather than resolved.

Together, these interactions form a stable avoidance equilibrium. No actor has both the incentive and authority to force resolution. Each can impose costs on others only by incurring significant costs itself. The services risk procurement paralysis if they reject systems outright. The bureaucracy risks political and institutional backlash if it forces cancellation. DPSUs risk funding and standing if they acknowledge failure too directly. Rational behaviour at each node therefore reinforces collective inertia.

This equilibrium is reinforced by limited external scrutiny and strategically manageable capability gaps. Secrecy fragments accountability, while emergencies such as Kargil or Galwan have been addressed through imports rather than structural reform. At the same time, the symbolic association of defence production with sovereignty raises the political cost of criticism.

The result is a system that does not resist failure but absorbs it. Each institution adapts within its own logic, and together they sustain a structure in which underperformance persists without triggering correction.

The AMCA Pivot: Bypass or Reform?

The decision to place the Advanced Medium Combat Aircraft (AMCA) program under a more private-sector-led structure marks the most consequential adjustment in India’s defence industrial system in decades.

Officially, the policy has been framed as an effort to widen the manufacturing base, increase competition, and accelerate innovation. Beneath that language lies a quieter conclusion. The Indian state no longer appears fully confident that the traditional DRDO-HAL model can independently deliver a fifth-generation fighter program within acceptable strategic timelines.

That conclusion has rarely been stated openly. Indian governments seldom acknowledge institutional failure directly in sectors tied to sovereignty, prestige, and technological nationalism. Instead, reform often comes indirectly through the creation of parallel structures that reduce reliance on existing institutions without openly challenging them.

The AMCA framework follows this pattern. Rather than fundamentally restructuring DRDO or imposing stronger external accountability on HAL, the state has begun constructing an alternative managerial pathway around the existing system.

In that sense, the AMCA initiative represents institutional bypass more than a reform.

The distinction is important because bypass and reform address different problems. Reform alters the incentives that produced underperformance in the first place. Bypass seeks improved outcomes without dismantling the structures that generated earlier failures.

The AMCA model may still improve program execution in important ways. Private-sector leadership introduces stronger schedule discipline, clearer managerial responsibility, and greater pressure to coordinate suppliers and development timelines within commercially meaningful constraints. Program credibility becomes more directly tied to measurable delivery.

These advantages should not be dismissed since India’s defence ecosystem has long suffered from diffusion of responsibility, where institutional prestige insulated program leadership from operational consequence. A privately led structure potentially narrows that insulation by tying credibility more closely to measurable delivery.

Yet the deeper accountability architecture identified earlier remains largely intact.

The clearest example lies in propulsion. The Gas Turbine Research Establishment remains central to India’s indigenous jet-engine effort despite the long history of the Kaveri program, which failed to meet the thrust and weight requirements necessary for operational fighter deployment. The evaluative structure governing engine development has changed little. Research progression still generates bureaucratic legitimacy even when deployable capability remains distant. A privately managed airframe program cannot fully compensate for a propulsion ecosystem still governed by internally mediated performance standards.

Nor does the AMCA framework fundamentally weaken HAL’s position. Even under a more decentralized model, HAL will likely remain central to production and systems integration once manufacturing begins. Its workforce networks, supplier ecosystem, and political protections therefore remain intact. The incentives that shield the institution have not disappeared. They have simply moved further down the production chain.

The Ministry of Defence also continues to act as intermediary between all relevant stakeholders. Information asymmetry therefore persists, along with the bureaucratic preference for compromise over direct institutional confrontation. The broader avoidance culture remains recognisable even if part of the system has been outsourced.

None of this makes the AMCA shift insignificant. On the contrary, the initiative may matter precisely because it reflects a more pragmatic understanding of institutional limits than earlier reform efforts. Rather than attempting politically costly restructuring, the state has chosen to reduce dependence selectively in areas of highest strategic urgency.

But bypass carries risks of its own. Systems that avoid accountability instead of institutionalising it often reproduce the same weaknesses in new organisational forms. If private-sector leadership succeeds, governments may continue building parallel structures around entrenched institutions without fixing the conditions that weakened them in the first place. If it fails, the existing public-sector system remains the default fallback.

In either case, the accountability problem survives.

The AMCA program therefore occupies an ambiguous place within India’s defence political economy. It reflects growing recognition that the older model has reached its limits in some strategic sectors. Whether that recognition eventually produces a different culture of accountability remains uncertain.

Policy Implications

India’s defence public sector endures not from policymakers’ ignorance of its weaknesses, but because the system distributes incentives such that accountability carries immediate, concentrated political and bureaucratic costs while underperformance remains strategically manageable. Effective reform must therefore work with these realities rather than against them. Abstract appeals to efficiency will be absorbed or stalled; targeted changes that raise the cost of avoidance without triggering the protective coalition offer a more viable path.

First, separate programme accountability from employment protection. The AMCA model of shortlisting private consortia (Tata Advanced Systems, L&T, and Bharat Forge) for prototype development in early 2026 while sidelining HAL from the initial phase illustrates this logic.

Program leadership should default to competitive private or hybrid consortia for major platforms, with DPSUs retaining manufacturing, systems integration, and sustainment roles where their scale and political footprint matter most. This decouples delivery responsibility (tied to schedules, costs, and performance) from local employment ecosystems that politicians across parties must protect. It allows employment continuity in key constituencies while exposing execution to commercial discipline and reputational risk.

Second, create external legibility of outcomes. DRDO’s internal metrics reliably reward research activity while operational deployment remains a distant, shared contingency. An Independent Program Assessment Office, reporting to the Defence Acquisition Council, should evaluate major programs against fixed benchmarks for schedule adherence, cost variance, and deployment feasibility. Its mandate would be narrow and evaluative, not executive: to shrink the space for indefinite metric substitution and “in-progress” status. Complementary reforms to DRDO (underway as part of the 2025 Year of Reforms) could reinforce this by sharpening its focus on core R&D while shifting routine development to industry.

Third, disrupt the avoidance equilibrium. The services, MoD bureaucracy, and DPSUs have settled into managed coexistence where open confrontation costs more than chronic delays. Grant the armed services statutory authority to trigger formal Defence Acquisition Council review when operational shortfalls breach predefined thresholds. Civilian and political authority would remain final, but the mechanism would end the current ability to absorb dissatisfaction indefinitely through informal accommodation and parallel imports. The evolving DAP 2026 framework, with stronger emphasis on Indian ownership of design and IP, provides a vehicle for embedding such procedural guardrails.

Fourth, establish minimal transparency. A public Program Outcome Registry should track original sanction parameters, revised timelines, cost variances, and milestone status for all major DRDO/DPSU programs. Technical specifications can stay classified; performance trajectories should not. Without systematic institutional memory, every delay appears exceptional and every workaround temporary. Transparency will not magically generate accountability, but durable norms rarely emerge in its absence.

These measures do not promise rapid transformation of a system built on decades of political accommodation. They aim instead for incremental shifts in the equilibrium: making sustained underperformance more visible, more procedurally costly, and harder to normalise. The AMCA precedent and ongoing DRDO adjustments suggest parts of the Indian state are already moving in this direction—quietly acknowledging limits without frontal confrontation. Building on that pragmatism offers the best prospect for genuine, politically sustainable change.

Conclusions

India’s defence public sector persists not because failure is invisible, but because the costs of correction are immediate while the costs of delay remain politically and strategically manageable. This has produced a stable equilibrium in which each institution adapts rationally within its own incentives, yet no actor bears the full burden of underperformance.

The armed services adjusted through selective induction and continued imports. The Ministry of Defence relied on procedural management rather than confrontation. DRDO and HAL learned to sustain programmes through reportable progress even when operational closure remained distant. Together, these adaptations preserved institutional continuity despite repeated shortfalls.

The broader implication is that democratic systems do not automatically discipline visible failure. Reform occurs only when the cost of preserving the status quo exceeds the cost of confrontation. India’s defence sector has not yet crossed that threshold. DPSU-linked constituencies make restructuring politically costly, while imports and foreign partnerships have kept capability gaps manageable. Secrecy and the symbolic prestige attached to indigenous technology have further insulated the system.

Technocratic reform proposals have repeatedly disappointed because they treat a political-incentive problem as an administrative one. The reforms proposed here therefore aim at something narrower: raising the cost of avoidance incrementally through greater metric visibility, external assessment, and formal review mechanisms.

The AMCA programme suggests that parts of the Indian state may already recognise the limits of the traditional DRDO-HAL model. Yet the shift remains a bypass rather than a full restructuring of the underlying incentive system. Whether these parallel pathways gradually weaken the existing equilibrium, or merely reproduce it in altered form, remains uncertain. Institutional change rarely begins with public acknowledgment. More often, it begins quietly, when established strategies no longer feel sustainable.

The views expressed are those of the authors and do not represent the views of CNAWS.

Image Source: ChatGPT