Over the course of the year, CTJ has published a series of analyses estimating the revenue impact of tax plans proposed by half a dozen Republican presidential candidates. It published a summary of the findings earlier today in a blog post.
These findings have sensibly drawn a lot of media attention, and the first question CTJ staff usually gets is “how do you come up with these numbers?” Here’s a quick answer to this question.
The starting point for Citizens for Tax Justice and the Institute on Taxation and Economic Policy’s work is the ITEP Microsimulation Model. This is a computer model built on a large database of more than 150,000 records, each representing a tax payer’s actual federal tax return. Since 1960, the Internal Revenue Service (IRS) has made these databases available to researchers seeking to understand the nation’s tax system and the impact of proposed changes.
The result is a researcher’s dream: mountains of data showing everything from basic data on wages and capital gains to details of itemized deductions and business tax breaks.
Starting with this database’s economic profile of the incomes of all Americans, the ITEP model can easily be used to estimate the effect of different tax rules. With a point and a click, the ITEP model can show the effect of, for example, increasing the top capital gains tax rate from 20 to 25 percent, or repealing the itemized deduction for charitable contributions on the taxes paid by every single one of these 150,000 records. The result is a statistically valid estimate of how much federal tax revenues would increase (or fall) as a result of such a tax change, which means we can use the model to generate revenue estimates on federal tax reform plans.
Because we know the income levels of every single one of these records, we can also use these results to show how tax changes would affect different income groups, from the poorest twenty percent to the very wealthiest 1 percent.
The ITEP model produces a traditional or “static” revenue estimate, meaning that in general it does not calculate how tax changes affect behavior as does “dynamic” modeling. The latter is a type of modeling advocated by those who support supply-side economic theories, which claim dramatic tax changes will spur economic growth. But as we have noted elsewhere, economists can’t agree on whether such an effect exists, which means that the most responsible approach to revenue estimating is to present a static analysis.
Taken on its own, the ITEP Model can only analyze changes in taxes that already exist at the federal level. This means that when we want to analyze an entirely new proposed tax at the federal level, such as a value-added tax or a national sales tax, we supplement the model using other data sources to augment the economic profile we get from the IRS’ income tax database.
We use a similar process to analyze taxes levied by state and local governments, such as property taxes and sales and excise taxes, that aren’t included in the IRS data.