In conclusion, subsidizing medical expenses through the tax code imposes significant economic burdens, reducing the net subsidy available to taxpayers. 10/

The implied costs for alternative subsidy structures simulated using the model estimates are slightly lower but still substantial relative to the overall tax savings, as tracking out-of-pocket medical spending likely accounts for a large part of the hassle.

9/

Finally, I estimate a simple model that makes use of differences in the value of the subsidy across states. The model generates an implied cost of claiming the subsidy ~80% as large as the tax savings, on average. For some groups, the implied cost exceeds the tax savings. 8/

What’s more – those who are eligible but don’t claim the subsidy actually have *higher* levels of medical spending than those who are eligible and do claim. They also have lower income, lower levels of education and worse measures of health. 6/

Using detailed data on income and out-of-pocket medical spending from the HRS, I estimate what people could deduct and compare them to what they actually do. I find about half of deductions are claimed, and incomplete take-up costs older households more than $5B annually in lost tax savings. 5/

Does anyone actually spend that much on medical spending? Turns out over $75B in IMDs were deducted in 2021, about 2/3 by taxpayers over 65. The forgone revenue from this deduction is almost $10B annually – 29% of the forgone revenue from the mortgage interest deduction! 4/

What you can deduct is quite expansive! (Caveat: I'm not an accountant and do not offer tax advice ☺️) 3/

First, what is this subsidy? If you spend more than 7.5% of AGI on out-of-pocket medical spending, the excess can be included in your itemized deductions. Your itemized medical deduction (IMD) reduces your tax liability by reducing your taxable income:

2/