TPC is out with a new research report, State Income Tax Expenditures, by Aravind Boddupalli, Frank Sammartino, and Eric Toder, that compares the tax expenditure budget (TEB) reports of three states, California, Massachsetts, and Minnesota, and the District of Columbia.
It’s a topic near and dear to my heart, on which I spent a fair amount of time during my career at House Research. See here and here. I promoted the idea (internally to legislators) and drafted the bill that required preparation of the Minnesota TEB back in the 1980s.
The TPC Report is well worth reading (only 12 pages long). It illustrates how variable the reporting of tax expenditures is from state to state, despite the fact that the authors picked four reports that are reputed to be among the higher quality of the state TEB reports. States (as well as analysts and Treasury and the Joint Committee staffs) don’t always agree on what a tax expenditure is. Disagreements at the state level are much common than at the federal level. Much of that is attributable disagreements about the baseline or reference tax structure. (I personally disagreed with a number of judgments that DOR has made in preparing its report. I many times have regretted my decision to parrot the very general terms of federal law in drafting the Minnesota statute and not providing DOR more statutory guidance as to what the reference or normal tax structure is. The House Research reports, as a result, deviated from some of the DOR decisions – in particular with regard to the treatment of the sales tax, which I think in principle is to be a broad tax on consumption, while DOR considers it to be a tax on final sales. That decision, for example, causes DOR to treat exemption of intermediate business inputs as tax expenditures, which the consumption tax norm would not.)
The TPC report also illustrates to me how ill-considered it was for Congress to repeal the ACIR. A federal government entity should be regularly doing work of this sort as part of its central mission and could, perhaps, promote some uniformity and consistency with greater authority and credibility. Instead, we’re left to rely on the good graces of the nonprofit sector (as good as TPC is and it is) for making these national comparisons of state practices and to make recommendations for improvements.
A few observations and nits to pick about the TPC Report and its coverage of Minnesota, since I’m pretty familiar with Minnesota’s practices.
Comparative percentages across states. I was modestly surprised that Minnesota had the highest percentage of its income and corporate income tax in tax expenditures – 61% versus a low of 51% for California. Unfortunately, the report dose not say exactly how those percentages were calculated, so I can’t get an impression (much less a clear answer) of the extent to which the differences are real or just a matter of how the reports define and estimate tax expenditures. Specifically, the Report does not indicate whether the adjustment made in its tables to eliminate a few of the very obvious reporting differences were used in calculating the percentages. Minnesota is the only state, for example, that treats the nontaxability of Medicare as a tax expenditure. Was that amount included in calculating the percentages? It is more than $500 million per year. The same goes for single sales apportionment ($379 million), which is excluded from TPC’s tables but is in the Minnesota report as a tax expenditure. Those are only a couple of the larger differences in reporting identified by the report.
As an aside, tax expenditure numbers cannot be summed to get an accurate total (even accepting the individual estimates as accurate), because they do not reflect the interactions if they were simultaneously eliminated. For example, some taxpayers would quickly revert to the standard deduction, while others would be pushed into higher tax brackets, and so forth. Moreover, behavioral responses to repeal are not taken into account at all (at least in Minnesota’s estimates). As a result, the percentage estimates are questionable as absolute numbers, but still could be useful as a comparative matter.
Treatment of Minnesota’s marriage credit. The Report lists Minnesota’s marriage credit as a tax expenditure. No surprise, given that the Minnesota TEB lists it as a tax expenditure. But the Report filters out filing status as a tax expenditure, because of differences among the states in how they treating filing status provisions. Is treating married couples or heads of households as a separate category of taxpayers with lower rates a tax expenditure? States come to different conclusions about that – Minnesota does not consider it a tax expenditure but other states do, so TPC did not included it in its tables. Minnesota’s quirky marriage credit is really an adjunct or alternative approach to rate reductions based on filing status. When Minnesota (in the 1990s) decided to reduce marriage penalties, it could have done so by simply increasing the married joint bracket widths to be twice the single brackets. That is the traditional way to eliminate a “marriage penalty” in the rate structure, but it also confers “marriage bonuses” on couples when one spouse earns all or most of the income. Minnesota chose instead to offer a credit to married couples when the lower earning spouse’s wages, self-employment income, or pension pushed the couple into a higher tax bracket than filing as two singles would. For some unknown reason, DOR treats that as a tax expenditure, but does not treat filing status differences as one. TPC followed that practice in the Minnesota table, but I would not count the marriage credit as a tax expenditure.
Minnesota’s waters edge unitary. The TPC Report (page 6) has a minor mistake in saying that Minnesota allows water’s edge as an election. Minnesota mandates water’s edge combined reporting, a practice that was upheld by the Minnesota Supreme Court (the US Supreme Court declined to review the case). Minnesota could still treat water’s edge apportionment (with worldwide unitary as the reference baseline) as a tax expenditure, but does not. As the TPC Report notes apportionment rules is a topic on which where there is no policy consensus on what the baseline or reference tax base should be.
The Minnesota numbers do not reflect the 2019 tax bill, as noted by the TPC report. That tax bill, because the major changes it made (going to TCJA’s higher standard deduction, its limits on itemized deductions for taxes and mortgage interest, a modest rate reduction, etc.) likely will make some material changes in the estimates in the 2020. There is no way TPC could deal with that but it will be interesting to see how much it changes the numbers – probably by quite a bit for a few provisions.
Differences in using direct versus tax expenditures. A constant issue for policy makers is whether to use a tax expenditure versus a direct expenditure for a new or existing program. The TPC Report is not intended to deal with that issue, but rather is about improved reporting and oversight of tax expenditures. However, it does obliquely engage the point with regard to film production incentives. Most states do that with tax credits. (The Report says that Georgia’s tax credit has an estimated cost of $500 million, which is shockingly high to me and I assume why the referred to it.) The Report correctly notes Minnesota’s on-again, off-again film production incentive is done as a direct expenditure (page 10’s reference to the Snowbate program). However, it passed up the chance to identify a much bigger difference among the reports – Minnesota has a very generous property tax refund program for both homeowners and renters. Minnesota’s FY 2020 expenditures on this program are estimated to be over $780 million. Because this program is administered separately, not as an income tax credit, the Minnesota budget treats it as a direct expenditure. (In the past, it was an income tax credit for renters and senior homeowners, but the state dropped that to realize one-time budget savings by paying/administering it separately which allowed deferring recognition of its cost until the next fiscal year.) Functionally, the program is little different than a tax credit, but it shows up in the regular budget. California’s, Massachusetts’, and DC’s programs are all tax expenditures because they are administered under their income taxes. The TPC Report is about tax expenditure reporting but failing to highlight this big difference seems like a missed opportunity. It clearly distorts financial reporting of spending (and Census reporting of state tax levels) slightly.
I do wonder whether differences in operating and budgeting for these programs – as direct expenditures by Minnesota and tax expenditures by California, Massachusetts, and D.C. – makes any difference in how much review and oversight they get by their respective legislatures. The conventional wisdom is that tax expenditures are hidden spending that is subject to much less scrutiny than direct spending. That seems like a subtext of the TPC Report and is an inference I generally agree with based on my many years working in the Minnesota legislature. But in the context of relatively high profile programs like renters credits and circuit breakers, I have my doubts. This general topic seems like a fruitful area for public finance economists to study. I have often thought the mixed practice of states in providing college savings incentives – both as tax deductions and credits for 529 plan contributions (tax expenditures) and matching 529 plan contributions (direct expenditures) would provide a good natural experiment to study, since there a multiple states offering both types of programs (in some cases both types in one state). Minnesota has alternated between both approaches over the last 20 years.
Bottom line: The TPC Report is a good start at comparing state tax expenditures and is worth reading. It certainly would be nice if states moved to more consistent and robust reporting, as recommended by the Report. The Report shows just how variable and weak that reporting is now even among states with supposedly very good reports. One baby step (not suggested by the report) might be for the financial accounting standards board to require reporting in CAFRs of the refundable amounts of tax expenditures. In my view, it is only through federal mandates (e.g., Census Bureau reporting) or FASB requirements that there will be any consistency allowing reliable cross state comparisons. If any such mandate were proposed for tax expenditures generally (i.e., beyond the refundable portion), the hue and cry of opposition from the states would be deafening and the complexity of writing the rules, mind boggling. Just doing something on the refundability portion would be a daunting task. Pity.
SALT Speak - Joel Michael's Blog 