In 2009, when Congress was debating the American Recovery and Reinvestment Act the Congressional Budget Office estimated that, if the stimulus became law, 8.4 million new jobs would be created by 2012 and unemployment would fall from 9 percent to 6.0 percent.

The American Recovery and Reinvestment Act did become law, and the federal government did spend almost $900 billion on solar panels, home weatherization, and bullet trains to nowhere, but none of the CBO’s economic predictions became true. Instead of falling by 3 points to 6 percent, unemployment fell by less than half that (1.2 points) to 7.8 percent. And instead of creating 8 million new jobs by 2012, the stimulus managed far less than that at just 2.6 million new jobs.

What went so wrong?

The short is answer is we don’t know because the Congressional Budget Office never showed the public the economic models they were using to make their economic forecasts.

This is simply not acceptable.

Congress does need a scorekeeper to provide budgetary estimates for the policy changes it considers. But at a bare minimum, that scorekeeper should be forced to show how its models work. Currently the CBO doesn't have to do that. It's a "black box," a secret formula even Congress can't be allowed to see, yet which the House and Senate must treat as if they were handed down on stone tablets at Mt. Sinai.

It's an indefensible situation.

That is why I introduced the CBO Show Your Work Act of 2019. This bill would require the CBO to publish its data, models, and all details of computation used in its cost analysis and scoring. CBO would keep its role as official scorekeeper of congressional budget proposals – but now the public and the economic community would be able to see what's going on in all those spreadsheets and algorithms.

This legislation would hold CBO to the same standard the American Economic Association's "Data Availability Policy" sets for all academic economists: requiring all paper authors to ensure their data "are readily available to any researcher for purposes of replication."

Policymakers need data and data analysis to do their jobs. But to do their jobs well, they need the best analysis. And centuries of practical experience tell us that transparency and replicability are essential to the pursuit and acquisition of knowledge. There is simply no serious argument for insulating the most influential economic modelers in the United States from the academic standards that govern everyone else.

We can do better as a Congress and a nation. We are never going to agree on what the best healthcare, tax, or energy policies should be. But when we make our arguments about the costs and benefits of our preferred policies, we should at least be willing to explain how and why our policies would work.

Making the CBO show its own work would be a great first step.