How Bangladesh engineered a power crisis it can no longer afford

Bangladesh's power sector has always been ambitious. It wanted to grow fast, build big, and finally silence the old jokes about load-shedding. In one sense, it succeeded: the country now has more installed capacity than it knows what to do with. In another sense, it has created a financial crisis so large that even the most creative accounting cannot hide it.

A comprehensive review by the government's National Review Committee (NRC) finds that the problem lies not in unavoidable market shocks or temporary crises, but in deep-seated governance and contractual failures. The report matters because it arrives at a moment when the costs of those failures have become impossible to ignore.

The mess we are in

The Bangladesh Power Development Board (BPDB) is now the epicentre of this crisis and the channel through which it spreads to the national budget and the wider economy. Its annual losses have exploded from Tk5,468 crore in FY15 to Tk50,565 crore in FY25, nearly a tenfold increase in a decade. Over the same period, government subsidies rose from Tk8,978 crore to Tk59,600 crore. These are not the numbers of a sector that is merely struggling; they are the numbers of a sector that has lost its financial anchor.

The core problem is brutally simple: BPDB buys electricity at prices far higher than what it sells it for. Bulk supply costs have reached Tk12.35 per kilowatt-hour, while the weighted average bulk tariff remains Tk6.63. Closing this gap would require tariffs to rise by roughly 86%, just for BPDB to break even.

If that happened, Bangladesh's industrial electricity prices would leapfrog those of Vietnam, China, India, and Pakistan by 80–90%. In a world where competitiveness is measured in cents per kilowatt-hour, such a shock would not merely hurt industry, it would risk pushing manufacturing elsewhere. Electricity would remain available, but increasingly unaffordable for the very sectors expected to drive growth.

Erroneous capacity

The irony is that Bangladesh did not arrive here because it lacked capacity. It arrived here because it built too much of the wrong kind, at the wrong price, under the wrong contracts.

Between 2009 and FY24, installed generation capacity increased more than fivefold. Yet system-wide utilisation has hovered around 40–50% in recent years. In other words, on a typical day, roughly half the system sits idle.

As of FY24, installed capacity exceeds system requirements by 5,600–6,700 megawatts even after accounting for prudent reserve margins. Under conservative assumptions, excess capacity still exceeds 3,000 megawatts. Add to this the 2,100–2,800 megawatts that cannot be dispatched due to fuel shortages, grid constraints, or logistical failures, and total idle capacity reaches 7,700–9,500 megawatts.

This idle capacity is not just a technical inefficiency, it is a fiscal liability. The annual carrying cost of stranded capacity, measured through capacity payments alone, is estimated at $0.9–1.5 billion. This excludes fuel pass-throughs, foreign-exchange adjustments, late-payment surcharges, and contingent liabilities. The true cost is higher.

Bangladesh has built a power system where the bills arrive even when the electricity does not, a feature it shares with gas.

Premium pricing

If overcapacity is one half of the crisis, overpricing is the other. A decade of contracts awarded without competition, without due diligence, and without alignment to fuel availability has produced a cost structure that is fundamentally unsustainable.

Across technologies, premiums are consistently and unusually high. Solar power purchase agreements show premiums of 70–80%, heavy fuel oil plants, 40–50%, unsolicited gas projects, around 45%. The Adani Godda import contract is a masterclass in cross-border generosity, paying roughly 50% more for the privilege of importing electricity that could have been generated domestically.

Add it all up and Bangladesh has achieved something extraordinary: a power system where the only thing more inflated than tariffs is the confidence with which the PPAs were signed.

For a country still climbing the lower-middle-income ladder, this represents a remarkable achievement in one domain, behaving like a petrostate without the petro. While citizens juggle rising living costs, the nation has quietly become a global pioneer in a new development model, committing to pay nearly $1–1.5 billion a year for electricity it does not even use.

These are not isolated mistakes made in good faith. They reflect a system that rewarded overpricing, encouraged unsolicited proposals, and normalised contracts that shifted risk to the public while guaranteeing profits to private parties.

The scale of rent

Seen in this light, Bangladesh's experience starts to resemble something more deliberate.

Bangladesh may still be a lower-middle-income country, but it embraced a first-world hobby under the past regime: overpaying for everything. And that, in its own absurd way, is a kind of economic miracle. Payments to power producers have increased eleven-fold since FY11, while actual electricity generation rose only four-fold. Capacity payments alone have increased nearly twenty-fold.

Taken together, the NRC's findings point to a conclusion that is uncomfortable precisely because it is so orderly. The fiscal strain now visible in BPDB's accounts is not the result of miscalculation or bad luck. It is the cumulative outcome of repeated decisions that prioritised contractual certainty for investors over affordability, efficiency, or system coherence. Once locked into long-term agreements, these costs became automatic, renewed each month, indifferent to demand, fuel availability, or economic context.

At that point, overpayment ceased to be an anomaly and became a feature. What began as emergency accommodation evolved into routine practice, and routine practice hardened into a model.

The rent extraction machine

The now-repealed Quick Enhancement of Electricity and Energy Supply (Special Provisions) Act enabled such contracts without competitive bidding. But the deeper issue is institutional, a political economy that incentivised rent extraction.

This conclusion follows from basic institutional logic rather than speculation about individual intent. The contracts in question were repeatedly approved over many years, across multiple projects, technologies, and sponsors, despite producing outcomes that were consistently fiscally damaging, commercially unbalanced, and far outside reasonable benchmarks. Occasional mistakes are possible in any system. Persistent replication of the same one-sided features, high guaranteed returns, weak risk sharing, limited competition, and insulation from downside risk, points instead to incentives embedded in the decision-making environment.

In such settings, rational actors respond predictably. Where large, durable rents are created through contract design, and where discretion is normalised and oversight weak, those rents tend to be contested, allocated, and shared through political financing, influence, revolving-door relationships, or informal patronage. The alternative explanation, that successive officials repeatedly approved contracts misaligned with the public interest without learning or correction, would require assuming a level of sustained institutional irrationality that strains credibility.

The more parsimonious explanation, consistent with the NRC's findings and BPDB's balance sheet, is that the system evolved to reward decisions that generated excess profits, and that these profits did not remain confined to the formal contract holders.

The NRC report makes this visible. What happens next will determine whether BPDB's losses continue to spill into the national budget and the wider economy, or whether this accounting clarity becomes the starting point for unwinding a system that has become too expensive for the country to afford. 

Zahid Hussain is a former chief economist of The World Bank, Dhaka Office