The Internal Revenue Service is being asked to provide more information about the race, ethnicity and sex of taxpayers but is facing legal restrictions, according to a new report.
The
In the report, the GAO found the tax data now collected by the IRS isn’t consistently linked to households’ demographic information. The Internal Revenue Service only collects the demographic data that’s explicitly referenced in the Tax Code. According to the Treasury Department, the IRS can’t collect demographic data under current law unless the information is necessary for administering the Tax Code.
“As a result, analysts have limited ability to assess the effects of tax laws, including COVID-19-related tax relief provisions, by demographics such as households' race, ethnicity and sex,” said the report. “Legal restrictions on interagency data sharing limit agencies' ability to analyze how the tax system interacts with households by demographic characteristics.”
Several agencies, such as the White House’s Office of Management and Budget, have emphasized the importance of collecting and sharing demographic data for policy evaluation, but they’ve also cautioned about the importance of protecting the privacy and security of those data.
The GAO identified several options for linking taxpayer and demographic data, such as surveys and interagency data matching. Another option is to “impute” the demographic information of taxpayers, which provides a kind of estimation. The Treasury is developing such an imputation method. For example, one imputation method that could be employed uses name and geographic data from the available administrative and survey data sources to determine the probability of a surname and geographic location being associated with a race. Those probabilities would then be used to correlate race information with data lacking that information.
However, there are drawbacks with using such methods. They can introduce errors in the missing data, and that could affect the reliability of summary statistics on tax outcomes by race, ethnicity and sex. Imputation methods can also introduce bias into the data, and that could lead to the wrong conclusions about the correlation between demographic factors and tax outcomes that may be inaccurate. For example, if analysts use income to impute race and ethnicity, then the correlations between race and tax outcomes may actually reflect a correlation with income, the variable used for imputation. Those limitations might become more pronounced when imputations are used to conduct detailed analyses of specific tax provisions.
While the Treasury is still evaluating the reliability and limitations of imputation, it hasn’t evaluated the feasibility of other options to produce data that would support more reliable analyses.
The report noted that the U.S. has a large and increasing gap in income and wealth by race, ethnicity and sex, but little is known about the effects of tax policies across demographic characteristics. While the Tax Code doesn’t tax individuals differently based on those kinds of demographics, some researchers have noted how it could result in potential unintended disparate tax outcomes.
There are some benefits to having better demographic data available on taxpayers. “If tax data could be linked to households' demographic data in a way that still protects the privacy and security of those data, policymakers and researchers would have better tools for consistently and systematically analyzing the relationship between tax policies and household demographics,” said the report.
The GAO recommended that Congress consider revising the relevant laws to facilitate interagency data sharing and also recommended that the Treasury evaluate the feasibility of other options to produce secure, linked taxpayer and demographic data. The Treasury said it is focusing on imputation and has considered other options. Moving forward, the GAO suggested that evaluating other options would enhance the Treasury's efforts to produce such data.
“Treasury’s Office of Tax Analysis is currently undertaking an unprecedented effort to analyze the relationship between tax policies and multiple demographic characteristics, including by improving on current statistical imputation methods to allow Treasury to model the relationship between race, ethnicity and taxes,” wrote Lily Batchelder, assistant secretary for tax policy at the Treasury Department, in response to the report. “This improved imputation method could be used with our various existing tax datasets and models, such as our individual tax and distribution models.”
Senate Finance Committee chairman Ron Wyden, D-Oregon, had requested the report from the GAO. “Policy is most often made through the Tax Code, but we don’t have concrete data on how the countless credits and deductions benefit different demographic groups,” he said in a statement Wednesday. “If we’re going to improve policy as it relates to child care, education and housing we need a clear picture. I thank GAO for their work on this report I requested, and am pleased to see their recommendation that we work to facilitate greater information sharing between the Treasury Department and Census Bureau. I’m committed to ensuring policy makers can obtain this greatly needed information from the Office of Tax Analysis and the Joint Committee on Taxation.”
Also on Wednesday, the House Ways and Means Oversight Subcommittee held a hearing on tax fairness in which they heard testimony from Kenneth Corbin, commissioner of the IRS’s Wage and Investment division and chief taxpayer experience officer, along with James McTigue, director of strategic issues at the GAO. They mainly discussed a separate
He noted that there were nearly 90 million visits to non-English pages on the IRS website IRS.gov last year. This year, through April 10, there have been about 17.7 million visits to non-English pages.
