The IRS is really obsessed with waiters and their tips. (Hey, you gotta make up for Hunter’s shortfall somewhere.)

In its Tax Gap studies, the IRS estimates that it gets 99% of what its due on regular wages, where taxes are withheld and reported to both the IRS and the taxpayer on a W-2, but just 55% of what it’s owed on tips (the same percentage it figures it collects from self-employed sole proprietors)…

Over the last three decades, the IRS has launched a series of programs that encourage employers to voluntarily calculate, report and collect taxes on a certain level of tips in exchange for protection from tip audits for themselves and their tipped employees. Except that’s not working great either. A 2018 study by the Treasury Inspector General for Tax Administration estimated that 30% of employers with tip agreements in place were underreporting. Tens of billions in annual taxes are at stake; TIGTA notes that the IRS itself estimated in 2006 that 10% of the individual tax gap comes from unreported tips.

Now the IRS wants to get in on the point-of-sale tipping action too. It hopes to modernize its data collection, and more effectively shift the burden for reporting tips to employers by using POS, time and attendance systems, and electronic payments data collected by employers.

Even while the IRS is cracking down on waiters, it’s engaging in the usual equity theater by accusing itself of racism.

Black taxpayers are audited at higher rates than other racial groups, an internal IRS investigation has confirmed.

“While there is a need for further research, our initial findings support the conclusion that black taxpayers may be audited at higher rates than would be expected given their share of the population,” IRS Commissioner Daniel Werfel told lawmakers.

In a letter to the Senate Finance Committee on Monday, Werfel said the agency would review its audit algorithms for specific anti-poverty tax credits to look for and address any racial biases.

“We are dedicating significant resources to quickly evaluating the extent to which IRS’s exam priorities and automated processes, and the data available to the IRS for use in exam selection, contribute to this disparity,” Werfel said in the letter.

Significant resources!

The actual paper this is based on has all sorts of issues with it, like not actually knowing if the people being audited are black.

the IRS does not systematically collect data on taxpayer race, either directly via tax returns or indirectly via merging tax data with administrative data on race from other agencies (Bearer-Friend, 2019). Instead, we rely on a BIFSG approach to estimate the probability that a taxpayer is Black (and non-Hispanic) based on the first name, last name, and location of the taxpayer.

So the IRS is apparently racially biased even though it doesn’t track race. If it doesn’t know if the people are black, then what is the basis for the bias? Racist algorithms, apparently.

Rather than measuring audit rates by races, this guesstimates whether someone is black based on their name and where they live, and then compares it to everyone else who is assumed to be non-black.

This isn’t how you produce any kind of reliable results.

Furthermore, it means that it isn’t comparing black people and everyone else, it compares black people who fit a certain kind of cliche, with certain kinds of statistically black names living in the ghetto. If you’re a middle-class black person living in a mixed area, you’re probably counted as non-black.