The Trump Administration has taken multiple steps that risk permanently damaging or eliminating important federal data sources. And senior current and former government officials as well as researchers worry that more serious threats are on the horizon, including to critical information that is key to informing tax policy.
Why should the public care? Because those data can help explain everything from burgeoning health risks to changes in weather, inflation, and employment. They provide important information about the tax system as well as the effectiveness of government programs such as education assistance, Medicare, and Social Security.
Part Of A Broader Effort
Suppression of tax data is not new. The Tax Policy Center was created more than two decades ago after the Treasury Department in both Democratic and Republican administrations began limiting the information it released on the distributional effects of tax proposals.
However, the magnitude of the threat to data collection and analysis may be unprecedented. In recent weeks the Administration has:
Effects On Tax Data
At IRS and Treasury, layoffs, resignations, and retirements will limit their capacity to collect and analyze tax data. According to published reports, the IRS already has fired about 7,000 employees, many others have resigned or retired, and the Administration plans to eliminate as many as half of all IRS jobs.
Already, the agency’s Research, Applied Analytics and Statistics Division lost more than 10 percent of its staff when probationary employees were fired in February. They included data scientists, economists, statisticians, and computer specialists.
Because of the magnitude of staffing cuts alone, projects will likely have to be abandoned. Cuts could come in programs such as Statistics of Income (SOI), which provides the most detailed information publicly available on, for example, overall deductions, credits, and sources of income.
Regardless of staffing, President Trump’s January 20 executive order terminating all federal programs related to “diversity, equity and inclusion” has halted IRS studies of how the revenue code and tax administration affect Black, Asian, and Hispanic households.
Future Worries
Based on what already has happened at other agencies, data experts tell me they worry that entire research programs at IRS and Treasury may be eliminated, specific studies may be stopped, or results suppressed.
Reports and studies that once were made widely available may no longer be open to the public. The Trump White House has a history of censuring data for political purposes.
For example, in the first Trump Administration Treasury Secretary Steven Mnuchin asserted a staff study found the 2017 Tax Cuts and Jobs Act (TCJA) would pay for itself but the study he claimed supported that conclusion was never released.
Looking ahead, will there be pressure to suppress statistics that show declines in customer service such as longer telephone wait times or reduced revenue collections due to staffing cuts? What about decreases in taxes paid by higher-income taxpayers, or increases in taxes paid by lower-income taxpayers?
What will happen to cooperative agreements between IRS and academic researchers? These agreements have produced scores of important research studies. Just one example: Harvard economist Raj Chetty and colleagues have written a series of groundbreaking papers on economic mobility, including the effects of income on health, based on IRS administrative data.
Reducing Efficiency
Over time, it may be possible to replace some IRS and Treasury staff with technology such as machine learning and artificial intelligence. But developing those models will itself require the work of staff data scientists, policy experts, and other researchers.
Slashing research capacity now risks making the federal tax system less transparent to outside analysts, lawmakers, and the public; stifling innovation; and leaving policymakers and tax administrators blind to the effects of what they do on households and businesses. All of which would make government less efficient.
Read the full article here
