When Governance Breaks Before the Dam

When a temporary cofferdam collapsed at the Tarbela-5 Extension Hydropower Project, the immediate explanations were predictable. Engineers debated design. Contractors defended construction methods. Commentators focused on delays, cost overruns and accountability. The images of floodwaters reclaiming the construction site seemed to tell the whole story.

They did not.

The collapse at Tarbela was not simply the failure of a temporary engineering structure. It was the visible end of a much longer process of institutional erosion. By the time concrete failed, governance had already failed.

That is how major public infrastructure seldom collapses. Not through a single catastrophic decision, but through a gradual weakening of institutional discipline. One procedural exception becomes acceptable. One approval receives less scrutiny than it should. One oversight mechanism becomes routine rather than rigorous. Documentation becomes less exacting. Responsibilities begin to overlap until accountability becomes impossible to locate.

By the time the public sees a collapsed structure, the real failure has already been unfolding for years.

Yet this description understates the problem. Governance does not drift by accident. It drifts because institutions face incentive structures that actively reward drift. The question is not whether procedures become flexible. The question is why those who manage the procedures have every reason to make them flexible.

Engineering projects are inherently uncertain. Large dams encounter unforeseen geological conditions, changing hydrology, contractor disputes, design refinements and construction challenges. No country can completely eliminate technical risk. The measure of a mature governance system is not whether problems arise but whether institutions identify them early, enforce discipline consistently and prevent manageable risks from becoming national failures.

Pakistan has repeatedly demonstrated that it possesses the engineering capability to undertake projects of extraordinary complexity. Tarbela itself remains one of the world’s great engineering achievements. The Tarbela-5 Extension represents an innovative effort to generate an additional 1,530 megawatts of renewable electricity by converting an existing irrigation tunnel into a hydropower facility. International financing was secured. Experienced consultants were engaged. Detailed contracts were prepared. Technical standards were established.

None of these, however, can compensate for weakening institutional discipline when that discipline is systematically undermined by the incentive structure within which institutions operate.

According to the government’s inquiry committee, responsibility extended across the three principal participants: the employer, the engineering consultant and the contractor. The inquiry identified unauthorised post-contract design modifications and failures to follow established approval procedures as contributing to the collapse.

If those findings are sustained, they tell us something profoundly important. Engineering knowledge was available. Procedures existed. Contracts defined responsibilities. Oversight mechanisms had been created. Yet governance failed to ensure that each part of the system performed the function for which it had been established. This failure was not inevitable. It was the predictable outcome of a system that rewards certain behaviours and penalises others.

Consider the structure of incentives facing each actor. The contractor completes projects faster through design shortcuts, reducing costs and accelerating payments from the employer. The consultant who raises objections to unauthorised changes risks project delays and client dissatisfaction. The employer who enforces strict compliance faces project postponements and political embarrassment. Each actor individually faces a rational calculation that non-enforcement is more advantageous than enforcement.

This is not incompetence. This is incentive alignment.

The same structure appears throughout Pakistan’s water sector with systematic regularity. Observe the irrigation system. Pakistan inherited and expanded one of the largest contiguous irrigation systems in the world, representing public assets worth approximately Rs 4 trillion. Its barrages, canals, regulators and distributaries continue to sustain the overwhelming majority of the country’s agricultural production. Yet much of this infrastructure has entered an age where preservation is becoming as important as expansion.

The engineering challenge is well understood. Infrastructure ages predictably and requires continuous maintenance: gates, equipment, linings, embankments all deteriorate over time.

The institutional reality is altogether different. A politician who inaugurates a new barrage wins constituents’ gratitude and press coverage. A politician who approves a maintenance budget wins neither. The political dividend arrives at ribbon cuttings, not at canal cleanouts. A bureaucrat who completes a megaproject advances in seniority and departmental prestige. A bureaucrat who maintains aging infrastructure stays in place. Departmental budgets grow through new projects, not through maintenance allocations. Consulting contracts for new design work generate revenue. Quality assurance on existing projects generates only cost.

The system is not broken. It is working precisely as designed. It selects for new construction over maintenance, for project speed over scrutiny, for approval convenience over independent verification, for approval speed over independent oversight. Governance drift is not accidental. It is structurally rewarded.

Inspection reports routinely identify structures requiring attention. Engineers understand which barrages need rehabilitation, which canal reaches are deteriorating and which regulators are approaching the limits of their design life. Maintenance schedules exist. Technical knowledge exists. What too often fails is not the institutional commitment to act. It is the political and bureaucratic incentive to act before deterioration becomes failure.

Preventive maintenance competes with immediate fiscal pressures. Rehabilitation projects are postponed not because engineers lack knowledge but because they produce fewer political dividends than new construction. Budget allocations fluctuate whilst infrastructure continues to age regardless of financial cycles. Deferred maintenance gradually becomes accepted practice until emergency rehabilitation becomes both unavoidable and vastly more expensive. Whilst this pattern unfolds in bureaucratic routine, farmers dependent on aging canal systems experience increasingly unreliable water supply that threatens harvests and livelihoods.

The same pattern that emerged at Tarbela is therefore not a deviation from normal practice. It is normal practice. It is what the system selects for.

Every governance failure carries an economic cost. The financial consequences of delayed hydropower projects extend far beyond construction budgets. Every year of delay postpones the generation of affordable renewable electricity. Thermal generation continues where hydropower should have replaced it. Fuel imports increase. Pressure on foreign exchange reserves intensifies. Electricity consumers ultimately bear higher costs whilst expected economic returns on public investment are deferred.

Similarly, every year that irrigation infrastructure deteriorates through inadequate maintenance increases future rehabilitation costs whilst reducing the efficiency of water delivery. Deferred maintenance means deferred economic growth. Poor governance is therefore not an abstract administrative weakness. It has a balance sheet. It has a fiscal cost measured in billions.

The response to this problem cannot be limited to procedural reform. Better procedures alone will not solve a problem created by misaligned incentives. Calling for stronger enforcement is, in effect, asking people to act against their own interests. No amount of governance training, audit frameworks or oversight boards will overcome the fact that the political system rewards new projects over maintenance, that bureaucratic advancement follows megaprojects rather than upkeep, that construction contracts generate more revenue than quality assurance.

Addressing this requires structural change to the incentive architecture itself. Departments must lock maintenance budgets into multi-year allocations rather than subject them to annual political discretion. Officials must have performance metrics that include asset preservation and operational efficiency, not merely new project delivery. Consulting contracts must include sustained performance guarantees that penalise early project abandonment and reward long-term asset management. Independent oversight must be institutionally separated from employer control, with authority and accountability located outside the project hierarchies that benefit from approval speed.

These are not calls for better implementation of existing procedures. They are calls for structural redesign of the incentive systems within which procedures operate.

Pakistan’s water crisis is often described as a shortage of water. Increasingly, it is a shortage of institutions whose incentives are aligned with the public interest. Good intentions cannot overcome a system that systematically rewards the opposite. Water security depends on institutions that can enforce discipline throughout an asset’s life. Without aligned incentives, neither discipline nor water security is possible.

The collapse at Tarbela should therefore be understood not as an accident or an isolated incident. It is evidence that Pakistan’s water system is working precisely as designed. The design simply does not reward the outcomes the nation requires.

Before Pakistan pours more concrete, it must rebuild the incentives that govern it. Otherwise, the next failure will not be an engineering surprise but an institutional certainty.