When tax legislation is proposed, the Arkansas legislature turns to the Executive Branch to get an estimate of how much additional tax money will be collected or how much will be lost through tax relief. The estimates are provided by the Department of Finance and Administration (DFA). These “revenue impact statements” are extremely important in determining whether tax legislation will pass or fail.
Yet in many instances, DFA may not have a clue whether their revenue estimates are accurate or wildly inaccurate.
In general, DFA has a good reputation for its work, but when it comes to revenue impact statements sometimes DFA’s estimates may be little more than flim-flam.
Whether accurate or wildly inaccurate, DFA’s revenue impact statements influence whether legislators vote “For” or “Against” a tax bill.
As an example, let’s look at Act 141 of 2017 which was touted as a military tax relief bill. (The act gave income tax relief to military retirees but other veterans were left out.) The legislation also included several tax increases including taxes on soft drinks, candy, and digital downloads (such as downloads of movies, books, etc.), and made unemployment compensation taxable as income under the Arkansas income tax. The legislation was touted as “revenue neutral” meaning the tax increases were supposed to offset tax revenue lost in giving special tax treatment to military retirees. (Even if it was revenue neutral, some taxpayers were winners and others were losers.)
It is now two years later. Let’s find out whether DFA’s revenue impact statement was accurate. Was the estimate of the cost of tax relief accurate? How accurate was the estimate of the new taxes? Was the overall bill revenue neutral or a tax increase?
Under the Freedom of Information Act, DFA was asked to provide:
- All records showing the tax revenue increase received under Act 141 of 2017 – a law that increased the taxes on soda, candy, and digital products.
- All records showing the tax revenue reduced from military retirement income under Act 141 of 2017 – a law that exempted income tax for military retirees.
Mr. Scott Hardin, Spokesperson/Communications Director of the Department of Finance and Administration replied:
“Regarding your first request, we can provide the general revenue reports, but gross receipts taxes are submitted by total sold by applicable vendors and the Department does not have details regarding the specific items sold. We can certainly provide the fiscal impact statements including our previous projections but the existing records of the Department do not track what the separate invoices are each month.”
In other words, DFA is saying they don’t know how much tax money they collected from these increases.
Regarding the second request Mr. Hardin replied:
“The Arkansas Department of Finance and Administration does not have responsive records that reflect the amount of tax that was not realized because of the exemption. Additionally, there are no calculations or reports that would provide this number.”
Turns out DFA has no idea how Act 141 of 2017 changed tax collections and DFA has no way of finding out. This means there is no way of judging whether DFA’s revenue impact statement had any merit at all. And, there is no way for DFA to know whether its subscriptions to computerized economic models provide useful estimates or just the basis for flim-flam.
Arkansas legislators have come to rely on the revenue impact statements provided by the Executive Branch (DFA). We doubt many legislators realized that sometimes these numbers have no basis in fact.
The question is: What will legislators do to ensure they have the best estimates possible and enough information to determine whether the estimates are reasonably accurate or wild guesses.
If you were on the House or Senate Committee on Revenue and Tax wouldn’t you be asking DFA serious questions, such as:
- What types of tax changes can DFA accurately track with hard data and what types of tax changes are hard to predict?
- What is DFA’s methodology were there is no hard data?
- Is the economic software used by DFA the most accurate available for predicting revenue impact and what systems are used in other states?
- Based on the type of tax change, what is the margin of error in revenue impact statements?
- What reasonable measures can be undertaken by DFA to track more actual results?
- What other sources are available to allow legislators to get a second opinion on revenue impacts?