Two Attorneys Claimed 944 Hours on Their Airbnb. The Tax Court Did the Math.

Two practicing attorneys against an IRS examiner. Who's your money on?

Most people would take the attorneys. These weren't taxpayers fumbling through Publication 925 — the Mirches ran a law firm in Reno. They understood burden of proof, evidence, and substantiation for a living. When they claimed material participation in their short-term rental, they backed it with a log showing 944.5 hours for the year — enough to clear every threshold in §469 with room to spare.

The court credited 7.4.

Not 7.4 hundred. Seven point four hours — out of nearly a thousand claimed. If you want to understand what actually wins these cases, that ratio is the whole story, because the difference between the hours that survived and the hours that died had nothing to do with legal skill. It had to do with what stood behind them.

The method was the problem

The problem was the method. Instead of recording actual work as it happened, the log assigned standardized time estimates to categories of tasks — the same number of hours for every occurrence, regardless of what actually occurred. The centerpiece was "site management": 8 hours for each of the 93 days the property was rented, defined as being "on call for guests, repairs, supplies, Wi-Fi, cable, snow removal." That single category was 744 hours — nearly 80% of the claim.

The court rejected all of it. Being available is not participating. The regulations require identifying the actual services performed and the approximate time spent performing them — and "I was on call" identifies neither. The court acknowledged the wife probably did answer guest questions and coordinate repairs during those stays. It didn't matter. Hours you can't tie to actual work performed are hours you don't have.

When your own return testifies against you

The cleaning category collapsed the same way. The log claimed 7 hours of cleaning per stay — every stay, whether the guest was there one night or fourteen. Meanwhile, the taxpayers' own return deducted nearly $10,000 in professional cleaning services, and they charged guests a separate $85 cleaning fee. Their own economic evidence testified against their time log. The court called the 168-hour cleaning claim a post-event ballpark estimate and credited essentially none of it.

The 7.4 hours that survived

Here's where the 7.4 came from: one category in the log did survive. The email hours — 12 minutes to read, 12 minutes to send — were found reasonable, because the court could check them against evidence showing the actual emails exchanged with prospective tenants. A grand total of 7.4 hours held up. The only hours that survived were the ones anchored to records that existed independently of the taxpayers' say-so.

Estimates versus evidence

That's the entire lesson of Mirch in one sentence: the court didn't grade the log on effort or plausibility — it graded each claimed hour on whether something verifiable stood behind it. Uniform estimates fail because real work doesn't happen in identical increments, and a log that says it does reads as reconstruction, not record. What survives is specific, contemporaneous, and corroborated: this call, this date, this duration, this vendor.

So the answer to the bet is that it was never attorneys v. IRS at all. It was estimates v. evidence, and evidence is undefeated.

Which is exactly what your phone already generates every time you coordinate a cleaner, chase a Wi-Fi outage, or schedule the snow removal the Mirch log could only gesture at. The work was real. The evidence wasn't. EvidenceGraph exists to close that gap — every management call recorded, transcribed, and time-stamped as it happens, so the record accrues while you work instead of being reverse-engineered when the letter arrives.

The Mirches had 944.5 hours on paper and about 7.4 in evidence. Don't let that ratio be yours.

---

EvidenceGraph provides documentation tools, not tax or legal advice. Consult your CPA or tax attorney regarding your specific situation.