Benzarti estimates the perceived compliance costs to tax filers by computing whether there were "too few" taxpayers with itemized deductions slightly greater than the standard deduction threshold. Taxpayers who might save a few dollars by itemizing might decide that itemizing was not worth the effort; those who could save much larger amounts would presumably incur the cost of keeping records and itemizing their deductions. Using Internal Revenue Service data from 1980 through 2005, the researcher found just such a pattern of "missing taxpayers" above the standard deduction threshold.
He next examined tax years in which there had been significant reforms in U.S. tax codes, specifically increases in the size of standard deductions. He focused on 1988, when the standard deductions were increased from $2,540 to $3,000 for single filers and $3,760 to $5,000 for joint filers, and compared that year's data to the years before and after 1988. He found that the level of deductions at which there were too few itemizers also shifted between 1987 and 1988, supporting the view that these "missing taxpayers" are the result of choices, not chance variation in the distribution of deductible expenses.
By estimating the amount of tax savings that each taxpayer forwent, Benzarti computed what they must have perceived as the filing cost of itemized deductions. He concluded that these costs are large, and that they are higher for high-income households. This is consistent with the opportunity cost of time being higher for these households than for others. Aggregate compliance costs appear to have risen over time, from $150 billion in 1984 to $200 billion in 2006 (both figures in 2016 USD). This suggests that compliance costs are about 1.2 percent of GDP in recent years.
Notice one large implication. Because the recent change in the tax law substantially raises the standard deduction, compliance costs should fall substantially. Imagine that they fell by 0.2 percent of GDP, that is, by 1/6 of the current cost. There's, in effect, a 0.2 percentage point boost. It won't show up in GDP: but it is what many of us would like GDP ideally to measure, that is, economic well-being.