Predictably, the liberal response to yesterday’s tax reform announcement by Senate leaders was the plan represents a big break to the rich and will shift the burden to the poor. Such criticism is very static in nature and uses very flawed methodology as its premise. I discussed the issue previously here and here, and also exposed a fatal flaw in the data used by a prominent left-wing group – challenging them to explain the gaping discrepancy in their data, which of course has gone unanswered for more than three months.
But Senate leaders may have given such critics a new tool to bolster their criticism in the form of the “tax calculator” on their website nctaxcut.com. It didn’t take long for the media to utilize the Senate’s own interactive tool to incorporate into their narrative.
For instance, the N&O wrote:
A married couple with two children making $30,000 a year would pay an estimated $1,000 more in taxes each year, according to a calculator on a political website designed to support the plan. By contrast, a single taxpayer making $200,000 would get a $6,000 break.
According to the to the tax calculator that Republicans rolled out to promote the reform effort, married couples with three children who make $40,000 would pay roughly $600 per year more in taxes every year. The same family earning $100,000 per year would see a cut of roughly $2,400 per year.
But how reliable is the data used to construct the calculator?
For instance, I entered in a household with $18,000 in income, married and no dependents. And because I, like everybody else, don’t know my annual expenditures with much accuracy, clicked on the “auto estimate” function for the calculator to automatically complete the information. The results showed that the new tax reform plan would actually result in an increase in the tax burden for this household.
But here is where it gets interesting.
The “estimated annual spending” total for this hypothetical married couple with no kids and $18,000 in income came to $17, 747. So I decided to enter the same information again, only this time increasing the dependents from zero to two. Lo and behold, the new “auto estimate” of annual expenditures for that hypothetical household was $17,742. So a household with two kids spent five dollars less than one with no kids, according to the calculator. Having two kids of my own, I know that ain’t right.
Just for kicks, I entered the same info in again, only this time increasing the dependents to five. This time the calculator estimated annual household spending to be $17,792 – only 45 bucks more than the household with no kids.
These results seemed off, so to investigate, I spoke to the developer at the Tax Foundation to learn more about this calculator’s methodology – especially in regards to how it calculates household spending for various income levels. The annual expenditures are based off of the Consumer Expenditure Survey produced by the Bureau of Labor Statatistics. Annual expenditures are an especially important component of the tax reform calculator because of the proposed expansion of the sales tax to apply to more household spending.
Looking more closely at this survey revealed the source of the problem.
For instance, take a look at this table summarizing a survey from last summer. Specifically, look under the column for the “lowest 20 percent” of income earners; the mean income after taxes in this column comes to $9,750. A little further down the column we find the average annual expenditures for this category to be $22,067.
In other words, according to the survey, the average low income household spends more than twice its after-tax income. What explains such a huge disparity?
A few possible explanations: low-income households underreport their income on the survey, overstate their annual expenditures, or count goods and services provided via government programs (i.e. food stamps, Medicaid, etc.).
In any event, because of this disparity – according to the Tax Foundation developer – the calculator places a “cap” on annual expenditures used for low income households to adjust, which helps explain why the above mentioned examples saw so little difference in expenditures between households with zero and five kids. More to the point, the calculator does the best it can at simulating results, but is limited by the data made available.
Another limitation of the calculator is that it doesn’t adjust for household expenditures consumed via government programs. In other words, the calculator will include sales taxes paid on the purchase of groceries under the new tax plan for all households – even those households buying groceries with food stamps, which is not taxed. Unfortunately, there is no data available to accomodate such an adjustment. With roughly 1.7 million North Carolinians receiving food stamps, that’s no small oversight.
The bottom line is this: the Senate tax calculator may be a useful tool for many users, but also can significantly skew numbers to make the reform plan look like it negatively impacts low income households, primarily because the available data sets are limited.