First, government needs to figure out what works. Beane’s successful approach was built on data and statistics — identifying those players who not only impressed with big swings, but also had a track record of successfully getting on base. We need a similar foundation for government programs. Right now, we know too little about the policy solutions that will improve outcomes, and we measure “success” by the numbers of people served or a handful of moving stories. Both President George W. Bush, and later President Barack Obama, took initial steps to gather more data and information about the impact of Federal investments. The Bush administration assessed approximately 1,000 programs for their effectiveness through the PART program. Obama built upon that effort and has prioritized budget requests that strengthen the use of evidence to improve education, train the workforce or reduce inequality. But rigorous evaluations of programs are still exceedingly rare, and Congress has not provided the resources – or the political will – to invest in building the kind of evidence that can be used to improve programs, make better funding choices and get better results. To fix this, the government should dedicate at least 1 percent of programmatic funding for evaluation, so we can begin to identify what works – and just as importantly, what doesn’t.

Second, once we know what works, government needs to shift dollars in that direction. When Beane looked at his small budget in 2002, he knew he had to spend his limited resources on the most effective players. The federal government, under any plausible path forward, will face increased fiscal constraints in the coming years. So it too needs to get the most for its dollars.