A new University of Michigan study finds that most missing self-employment income in tax records among older workers is due to unreported earnings rather than differences in income classification.
The study shows that self-employment income reported in the Health and Retirement Study (HRS) but missing from tax data aligns with IRS estimates of unreported earnings.
Mind the gap
The IRS estimates the net tax gap—the difference between owed and collected taxes—was $606 billion for 2022, based on random audits.
Researchers analyzed HRS data from 2004 to 2016 and found significantly more self-employment income than tax records show, supporting IRS compliance estimates.
“While the IRS uses random audits, our research relies on survey data from HRS,” they explain. “It provides insight into who underreports self-employment income, how much is missing, and the types of work involved.”
The findings challenge previous research suggesting people are less likely to report self-employment income in surveys than in tax data.
A growing trend
“Most missing self-employment income is not due to misreporting but rather earnings that are not reported to the IRS,” the researchers state. “People engaged in informal self-employment are less likely to report income to tax authorities but disclose it in surveys.”
Underreported income is most common among those reporting very low or very high earnings. Those earning between the 5th and 75th percentiles underreported an average of $4,000. In contrast, those below the 5th percentile underreported $21,000 on average, while those above the 95th percentile underreported $27,000. The latter is a conservative estimate since earnings above $250,000 are masked in the data.
Unreported income limits government revenue, affecting budget deficits and funding for public programs. These findings could help improve audit targeting. They also highlight flaws in tax and survey data, which inform policies and regulations.
Missing self-employment income skews estimates of poverty and inequality, affecting well-being assessments and policy decisions. Ongoing research examines how reporting discrepancies vary by race and sex and how different income measures impact poverty estimates. Researchers also explore how various forms of self-employment influence well-being.
