Consider for a moment the task of an examiner at the Office of Management and Budget, one of many civil servants responsible for compiling President Obama’s budget for the 2012 fiscal year. This examiner has been cooped up in the New Executive Office Building since mid-September, hashing through agency budget requests and identifying which programs are worthy of federal funding.
Each program’s proposal presents a compelling case for why it should be funded. But with every program’s proposal boasting the best possible information on the program, however, how is our examiner meant to work out what is genuinely a good idea and what is not? The reality is that OMB examiners have dangerously few resources at their disposal to sort robust evidence from less compelling claims of program effectiveness.
But there is a way the federal government could deal with this problem effectively and efficiently. In a recent report, “Scaling New Heights,” CAP argued for the creation of “Institutes for Effective Innovations.” These small, independent entities—comprised of experienced researchers and experts in a specific policy area—would help to sort the good ideas from the bad. They would marshal evidence of what works in their field and provide an unbiased assessment of the effectiveness of individual approaches.
These new Institutions for Effective Innovations would not generate data for themselves, but collate all available information and assign an overall score. As a simple example, if an approach toward, say, reducing local poverty rates had been shown to work in 10 different states based on a series of randomized control trials, it would receive a high score. But if the approach had been tried in a number of places and the results were mixed, it would get a lower score. And where there was no evidence that the approach had ever worked it would get an even lower score.
So by the time the proposal hits a desk at OMB as one of hundreds of new programs competing for funding, the score would help with the sorting process. Facing strict resource limitations for new programs, programs with low scores would find minimal support.
For more on this topic please see: