The Gates Foundation has pledged $40 million to independent think tanks in developing countries, starting with a 24 institutions in Africa.

The aim of the initiative is to provide long-term funding to organizations so they can produce sound research that influences national policy debate and decision making, said Mark Suzman, director of policy and advocacy for the Gates Foundation's global development program.

The Gates Foundation and two partners, Canada's International Development Research Centre and the William and Flora Hewlett Foundation, pledged a combined $90 million in grants to the effort.

"Effective development in a sustainable way really only happens when you have committed national governments putting in place the right policies based on the right information..." Suzman said. "There's a limit to what outsiders can do even with the right intentions."

The grants will go to organizations focusing on economics, technology, social and environmental policy, in countries including Benin, Burkina Faso, Ghana, Ethiopia and Senegal. Four think tanks were funded in Kenya and Nigeria and three in Tanzania. A complete list is here.

Although the funding is unrestricted, Suzman said the think tanks are expected to do research in many of the areas where the Gates Foundation works.

"We do hope and expect much of the work they do will support areas we at the foundation care a lot about, like health, agriculture and financial services," he said.

One focus of the research will be agricultural development.

"We know that African countries nominally committed to spend 10 percent of their national budgets on agricultural issues," Suzman said. "The question is what is the best way to spend those resources. The answers need to be locally based."

The think tanks receiving grants range from very small and new, with only a couple of employees, to older, more established institutions. More than 300 submitted applications. They are required to be independent of government with no links to a political party and a track record of peer-reviewed, evidence-based research, Suzman said.

But some of their work will involve gathering the most basic data.

"It's important not to underestimate just how little knowledge is available in some countries of things we take for granted," he said. "What is the real distribution of poverty, what are the regions where things are working well or not? Often they just don't have that information at hand."