Who decides what gets rebuilt first

Reconstruction is usually discussed as a financing question. It is at least as much a decision-making question: with more needs than money, somebody chooses, and how that choosing is organised determines both the outcome and whether people accept it.

The three sets of priorities

National: strategic infrastructure, energy, transport corridors, and the sectors a government judges to matter for the country as a whole.

Community: what a specific town needs, proposed by the people who live there and know which school and which pipe.

Donor: what a financing partner is willing to fund, which reflects its own mandate, its parliament and its sectoral expertise.

These overlap substantially and not completely, and the gaps are where the difficulty lives.

What a good process looks like

Published criteria: how projects are scored, weighted and ranked, stated before the decisions rather than after.

A pipeline that communities feed into, so a proposal has a route rather than depending on who knows whom.

Coordination between donors, so that four partners do not fund four schools in the same district while another has none.

And publication of the decisions with reasons, which is the only thing that makes an unfavourable outcome acceptable to the community that lost.

The failure mode to avoid

Allocation by relationship, where a well-connected mayor secures funding and a competent but unconnected one does not. That is how a recovery programme loses legitimacy, and it does so quietly and irreversibly.

The test

Can a community that was refused read why. If not, the process is not governance; it is discretion with a secretariat.

In the room where three sets of priorities collide, what wins is not the best argument but the most complete file. I see this in tenders every day: the job with drawings done, title clean and permits in hand goes first, because it is easy to give money to. The most useful help you could give communities is not money but support in preparing files, and that is never discussed.