KALAHI-CIDSS let Philippine villages fund, rank and build their own projects

The Philippines decentralized in 1991, but the money and planning mostly stopped at town hall: local governments had little practice sharing budgets with villages, and poor barangays rarely saw projects matched to what they needed. In 2002 the government's anti-poverty program took another route: KALAHI-CIDSS, run by the Department of Social Welfare and Development with World Bank support, targeted the poorest quarter of municipalities in the 42 poorest provinces — 200 municipalities chosen by poverty index and validated on the ground.

The design gave communities the budget and the ballot at once. Each municipality received a formula allocation of 300,000 pesos (about $6,667) per barangay, and barangays competed for it by proposing sub-projects and scoring each other's in open Municipal Inter-Barangay Forums; villages that lost one cycle gained priority in the next. Winning projects — roads by far the largest share, then water systems, schools, day-care centers — were built by volunteer committees with municipal technical support, under posted transparency requirements, third-party monitoring and a grievance mechanism.

It delivered at scale. Across the original financing, the additional financing and an Australian grant, the program funded 8,607 sub-projects in 6,144 barangays reaching 2 million households, with 75,700 short-term jobs and $27 million in wages paid. An impact evaluation across 2,400 households found per-capita consumption about 12% higher, funds reaching the poorest villages, and no elite capture. Community-built sub-projects cost 15–20% less than line-ministry equivalents, and the completion report rated outcomes, Bank performance and borrower performance all satisfactory.

Villages got the money — and each other's votes

The formula removed the discretion that turns grants into patronage: mayors could not re-route money, only advocate for proposals.

Peer ranking by the people who would live next to the results beat remote technical scoring — villages knew which water point would actually be maintained.

Elected volunteer committees doing procurement and construction built the skills and the stake; 1,005 trained volunteers carried the practice into other local work.

Transparent posting plus NGO monitoring plus a grievance channel made capture visible, which is partly why the evaluation found none.

Money without a voice becomes patronage; voice without money, theater. A formula fixes quantity, peer ranking fixes priority, and posted budgets fix leakage.

The additional financing in 2010 ($59 million) and Australia's DFAT grant in 2012 carried the model through. In 2014 the government launched the National Community-Driven Development Program to take the KALAHI-CIDSS modality to more than 900 municipalities nationwide — the scale-up the completion report said the project had deliberately laid the foundation for. KC-trained volunteers and systems were also folded into post-Yolanda reconstruction, testing the model under disaster conditions.