Congress and President Trump are embarking on what is likely to be a major rewrite of the federal income tax code. Yet, neither they nor anyone else knows whether the hundreds of tax preferences embedded in the law accomplish their stated purposes.
Federal agencies routinely collect and assess evidence of the success—or failure—of spending programs. But the IRS is an outlier. Although it manages a broad range of federal programs through the tax code, including housing, health, wage subsidies, and retirement, it has made little effort to evaluate its portfolio.
These policies, sometimes called tax expenditures, add up to more than $1 trillion per year and comprise approximately one-quarter to one-third of combined spending and tax subsidies. Yet the IRS appears to do less than any other agency to gather evidence on the efficacy of the programs it runs.
Years ago, the Office of Management and Budget instructed executive branch agencies to measure performance. Many did so. But the IRS essentially said it didn’t have the information, and did not comply.
Over the years, government has tried repeatedly to gather solid evidence to justify new and existing government policies. Success has been limited. It has tried “Planning, Programming, and Budgeting,” and “Zero-Based Budgeting.” Congress enacted the Government Performance and Review Act (GPRA). Speaker of the House Paul Ryan (R-WI) and Senator Patty Murphy (D-WA) advanced this goal recently through the creation of the Commission on Evidence-Based Policymaking.
The National Academies of Science has appointed committees to investigate both how to improve the quality of economic evidence and encourage its use to inform policy. I was fortunate enough to chair one of those committees. Jason Furman, President Obama’s last chair of the Council of Economic Advisers, recently noted how new evidence advances our ability to design social policy.
Overall, success has been mixed. But rarely have we learned less than at the IRS. My purpose is not to assess blame. Congress has given the IRS far more responsibility than it can possibly manage, even as it has slashed its budget. That is not likely to change any time soon. But even in that environment, there are ways the IRS can develop and present economic evidence on the programs it runs.
I have two simple suggestions. Both require the Commissioner to place greater priority on generating evidence.
First, the IRS should report administrative information to Congress on each program or subprogram it operates. It should describe how tax subsidies are distributed by taxpayer income and geographical location. It should report on noncompliance or on gaps in its ability to enforce the law in each program. To assess a program adequately, we need solid evidence on how well IRS can administer it.
Beyond basic administrative information, the agency could at least on a rotating basis start to evaluate program benefits and costs.
Second, the IRS commissioner should hire a Chief Evaluation Officer who would work closely with the Taxpayer Advocate to teach and encourage staff to make better use of the tools that generate such evidence. Demetra Nightingale, who recently served in this capacity for the Labor Department, says these tools were most successful when staffers believed they could help them do their jobs better, and not simply the result of some top-down command.
The most successful of modern reforms, the Tax Reform Act of 1986, succeeded in eliminating some tax breaks and deterring some shelters. But even it added new complexity to the code. If Congress is going to look seriously at the scores of subsidies in the current code, it needs better evidence about which achieve their goals and which can be properly administered. To do that, Congress needs to give the agency the resources it needs to properly evaluate the programs it manages. And the agency needs to do more analysis with whatever resources it has.
