| I |
n the summer of 2022, Pakistan was engulfed by its worst floods in living memory. A third of the country was underwater, affecting 33 million people. The United Nations’ post-disaster needs assessment (PDNA) estimated the damage at $14.9 billion, losses at $15.2 billion and the cost of resilient recovery at $16.3 billion. Sindh bore nearly three-quarters of the burden and Balochistan accounted for another fifth. Housing, agriculture and transport topped the damage tables.
Bilateral and multilateral development partners were quick to respond. At a conference held in Geneva in January 2023, donors pledged more than $10 billion for reconstruction under the Resilient Recovery, Rehabilitation and Reconstruction Framework. The headline numbers looked generous. Three years on, with Pakistan once again under water, the financing story has proved far less reassuring.
The composition of Geneva’s pledges tells the tale. The Islamic Development Bank offered $4.2 billion; the World Bank $2 billion; the Asian Development Bank $1.5 billion; the Asian Infrastructure Investment Bank $1 billion; and Paris Club countries $0.77 billion. Most of this came in the form of loans and project finance; grants were scarce ($0.54 billion). The imbalance matters a lot for a country already servicing debt with more than half of its federal revenues.
Another aspect of those pledges is that 58 percent of this amount is project financing, whereas 42 percent is oil financing.
By June 30, 2025, a large share of pledges had been converted into commitments and approvals; approximately $6 billion in approved projects and $1.9 billion in oil financing.
Flagship projects included a $450 million World Bank package for resilient housing in Sindh approved in December 2024; an ADB $400 million concessional loan for housing and community infrastructure approved in July 2024; and an IsDB envelope of around $200 million for co-financing. Together, these targeted up to 770,000 homes with flood-resilient core designs.
The challenge is not that the Geneva pledges will increase our debt burden. The issue is that some of the flows have been re-labelled: commodity financing, budget support or deposits at the central bank rather than direct reconstruction funds. The real issue is that millions of people still live in tents.
I gave this lengthy preamble to set the scene for this year’s floods. The 2025 monsoon is still ongoing so that it is too early to assess the loss and damage. However, if 2022 is the benchmark, the eventual bill will again be counted in tens of billions. Furthermore, considering the quality and speed of disbursement of external financing for floods in 2022, we should temper our expectations from external financiers.
Another lesson from the 2022 floods is that we should now be quite familiar with what works and what does not. The federal government negotiates external resources; the provinces deliver on the ground; and the local governments (if present) are the best assessors of losses. However, the three, often don’t connect. On the other hand, even when concessional, debt-financed recovery compounds fiscal stress. Multilateral development bank projects move slowly by design, with feasibility studies and safeguards. Housing schemes are easier to replicate but transport, irrigation and schools lag behind. Reconstruction has favoured titled landowners over sharecroppers, formal settlements over informal ones. Grants for tenants or labourers have been sporadic. Projects that are highly visible, such as housing blocks and paved roads, attract financing faster than those that are less photogenic but equally vital, such as drainage rehabilitation or fodder support.
Floods are no longer once-in-a-generation event. With glaciers melting, monsoons becoming more erratic, and India’s lack of cooperation on vital data exchange, they are recurring fiscal shocks.
To break the trap, we need to rethink how flood finance is raised, structured and spent. More grants, contingent credit lines and insurance pay-outs are essential. Catastrophe bonds or regional risk pools could offer quicker liquidity than donor conferences. IMF financing should complement, not replace, grant flows. The National Disaster Risk Management Fund should be capitalised as a first responder, with regular budgetary allocations, possibly financed by a climate levy. Provinces should earmark part of their revenue share under the National Finance Commission award for resilience and disaster management.
We should also develop rolling damage assessment systems using satellites, drones and digital registries, publishing fortnightly updates to enable development partners to disburse funds in tranches instead of waiting months for a final PDNA. Development partners should publish disbursement dashboards that clearly distinguish between grants, loans and repurposed budget support, narrowing the perception gap between pledges and reality.
Compensation frameworks must explicitly include tenants, sharecroppers and informal settlements. Reconstruction must price resilience into every rupee, from elevated, flood-resistant housing to re-engineered culverts and stronger schools and clinics. Every flood exposes Pakistan’s structural weaknesses; recovery packages should therefore incentivise initiatives that reduce vulnerability in the long term.
Pakistan cannot borrow its way out of climate vulnerability, nor can development partners keep recycling loans as if they were relief. Development partners should work with Pakistan by providing contingent instruments that automatically release funds when specific thresholds are crossed. If both sides act, floods will still batter lives and livelihoods, but they need not paralyse the economy every time. Without such pragmatism, we will remain trapped between rising waters and rising debt.
The writer heads the Sustainable Development Policy Institute and is a member of the Asian Development Bank Institute’s Advisory Board. His LinkedIn handle is Abidsuleri.