Category: Uncategorized

Minnesota Management and Budget (MMB) is out with its second quarter report on state revenues and it’s breathtaking  – $2.7 billion above the amount forecast in February. That amount, of course, was well above what was forecast in November. Revenues are up across the board – income, sales, corporate, and every other tax, big or small.

Corporate revenues are up by the highest percentage (over one-third), although they comprise a small share of the total in dollars. The pandemic has been remarkably good to corporations, as evidenced by FRED’s pretax corporate profits series, which shows after a one-quarter dip (the shutdown 2^nd quarter 2020) profits quickly rising to all-time highs.

I don’t think TCJA related changes in reporting or how profits are booked has much if any role in the rise, but I suppose it’s possible. The buckets of money the feds have shoveled into the business sector (aid for the airline and hospitality sectors in addition to general aid like PPP loans) is undoubtedly responsible for much of this, along with the COVID relief aid, economic stimulus, and the resulting recovery.

If you had asked me to assign a probability to this revenue scenario in late spring 2020 (i.e., when MMB did its first post-pandemic forecast), I would have assigned it a trivially low number (5% tops). Of course, I never thought the federal government would pour the trillions of dollars of pandemic assistance and stimulus into the economy that it has. It was, in some ways, a perfect alignment of political planets – (1) the pandemic shock induced Congress to put aside partisan divisions in March and April and enact generous aid particularly the CARES Act, (2) the Republicans desire to give Trump a good shot at reelection allowed them to ante up even more in the fall despite their general antigovernmental posture (Dems are never reluctant it seems, even though it might not be in their short term partisan electoral interests), and (3) the seeming miraculous outcome in the Georgia senate run-offs that allowed total Democratic control and a third generous dollop in early 2021.

The economic effects so far make clear the power of federal stimulus – particularly, I would posit when the money is allocated as new spending or aid and tax credits to lower income folks (Dems’ preferred approach), rather than as traditional tax cuts (GOP preference). The former is more effective at stimulating demand, while the latter mainly drives up asset prices (stocks, bonds, etc.). If you give money to low-income consumers, businesses must invest and increase output to get their hands on that money. If you cut corporate taxes, they can take the easy way out and increase dividends or stock buy backs. If you cut the taxes of upper income folks who already have largely satisfied their consumption preferences, they will simply use the dough to bid up stock and bond prices or otherwise save much of it to assuage any economic insecurities they may have. That’s my simplistic view anyway.

So, when the federal stimulus booster rocket stages fall off and the economy must make it on its own will the trajectory of revenue growth also return to earth? I have no idea, but MMB forecasting methods, although cautious and generally conservative, are partially driven by baseline revenues. That likely will provide some momentum bias in the next couple of forecasts. Put another way, turning points, if that is what were in for, are notoriously difficult to call. Of course, I tend to be a Cassandra, born of way too many deficit budget sessions caused by an inherent tendency to over-optimism during good times. Caveat emptor.

Budget risks

The reality is revenues will not continue to come in at this pace. But even if they slow down a lot, the governor and 2022 legislature will face a big budget surplus – potentially of unprecedented size. Making matters worse, it will be a year in which everyone (governor and all legislators) is on the ballot, thanks to redistricting. The urge to spend (Democrats) and/or cut taxes (Republicans) will be irresistible. If recent sessions (2019 and 2021; the pandemic made 2020 an outlier) are any guide, the natural path/compromise is to do some or a lot of both. I suppose an alternative is partisan deadlock with little happening as both parties bet that the election will increase their leverage in 2023 and, thus, their need to compromise their agendas. Hard to know what to predict.

Because some unknown portion of the revenue growth – potentially a lot – is one-time, permanently increasing spending or reducing revenues would be highly risky. But one-time spending or tax cuts are typically unappealing politically to both parties’ bases and legislators think it doesn’t really move the ball up the field, so to speak. To state the obvious, the flood of onetime federal aid has already has already satiated much of the politically appealing possibilities. (Also, the own source revenue growth may limit options for using the federal aid for tax cuts under the complicated federal rules.) Hence, the risk.

In my view, a good use of the money would be to clear the state’s tax books of one-time items by, for example:

• Immediately allowing deduction of the disallowed amounts of section 179 allowances for tax years before full conformity applied – say in equal amounts in tax years 2022 or/and 2023.
• Reversing the June accelerated sales tax payments so that the same rules apply to all months.
• Increasing the percentage of bonus depreciation allowed in the year the property is placed in service and allowing faster deduction of previously disallowed amounts.
• Adopting other federal nonconformity timing items.

All these changes would have one-time or heavily front-loaded effects and would free up resources in the future. The problem is that their politics are unappealing, heading into an election. Businesses and Republicans won’t view them as real tax cuts, since they will regard the changes as just timing changes and simplifications, desirable to be sure but nothing to get excited about. The Democrats obviously won’t want to do them because they do not move their policy agenda. Both parties will view them as taking money off the table for the 2023 session (compared to doing nothing), which will be undesirable in their minds. However, they would preserve fiscal flexibility and could be used to tamp down unreasonable public expectations (i.e., the size of future “surpluses”).

Bottom line: a large budget surplus will make it politically challenging for legislators to avoid digging a budget hole that the fiscally prudent may later regret. It’s movie I seen too many times.

DISCLOSURE: I was one of the drafters of the 2012 stadium legislation. As a result, I am not an unbiased source. My views are clouded by the perceptions I developed during that process and working on other pro sports facility finance bills going back to the bill that financed the Metrodome in my first legislative session (I had a minor role working on its bonding provisions).

The reserve account serves no legal or financial purpose.

It is useful to briefly recover this ground (skip this if you read the first post). The legal security for appropriation bonds is only the statutory general fund appropriation. The practical or financial security for the bonds is the stain on the state’s credit that would result from the legislature repealing or modifying that appropriation and the massive resources in the state general fund. Wall Street, as a result, is confident the bonds will be paid. Thus, a reserve is not needed to provide assurance to the bondholders. Second, there is no dedicated stream of revenues to put into an account. Bond counsel advised that making the account look or function like a real reserve could undermine the bonds’ legality as appropriation bonds. That required drafting the law to cloak the account and its calculations in ambiguity (relative to the bond obligations). Thus, the account is a purely a political construct.

That reality, of course, does not mean retaining the account might not serve a useful purpose. This post addresses that possibility. To be fair, some of my criticisms are fundamentally criticisms of the underlying arguments for the original stadium deal. My point, as outlined in the second post, is that those arguments were crucial to selling the deal. Now that the stadium is built, the Vikings are here to stay (for a while anyway), and the main political players have left the scene, it is no longer necessary to continue the charade or Potemkin Village façade – i.e., that it was financed with incremental revenues generated by e-gaming, rather than by diverting preexisting general fund revenues. The reserve is a feature of that façade with undesirable side effects – unnecessarily reserving an ongoing stream of lawful gambling tax revenues from the standard state budgeting process. Put another way, its calculations increase the effective cost of the stadium each year that money is sequestered in the reserve. That now occurs annually.

The account’s flaws

#1: Available revenues in the account’s calculations are neither (1) generated by or related to the stadium nor (2) an accurate measure of incremental or increased revenues caused by the stadium bill.

The first point – i.e., that the revenues are not user or benefit charges related to the stadium – is so clear it needs little explanation. Lawful gambling taxes and the Minneapolis sales tax revenues do not derive from or relate to stadium uses and/or benefits (well, maybe a small part of the Minneapolis sales tax does because the stadium channels more economic activity into the city – not worth considering).

The second point – that the account’s revenues seriously overcount the incremental revenues that resulted from enacting the stadium bill – needs elaboration. The bill’s supposedly generated incremental revenues by authorizing electronic lawful gaming (e-pull tabs mainly). Doing will generate some additional state revenue (taking all economic considerations into account), but the account’s calculations do not make a serious effort to measure those revenues. Instead, its calculations count revenue – very likely a lot – that the state would have received had the stadium bill never been enacted. That is so for several reasons:

• The calculations assume the increase in all lawful gambling tax revenues over the 2012 forecast ($36.9 million/year) are new or “available” revenues. These amounts are not limited to taxes from e-games. Even if they were limited to revenue from e-games, some of that likely results from players of paper games switching to e-games. The fiscal note assumed traditional (paper) game sales would drop by about $200 million per year (about one-sixth) because of the availability of e-games. That was probably little more than a guess.
• The baseline was set at the end of the Great Recession, an artificial low point. For example, in fiscal year 2000 lawful gambling tax revenues were $62.5 million or more than 60% higher than the $36.9 million baseline in the account’s calculations (average annual revenues over the previous decade were 17% higher than the baseline). As the economy recovered, revenues would too, regardless of authorizing e-games.
• Much of the spending on e-games likely is diverted from other entertainment spending that pays sales tax (e.g., bar and restaurant purchases, admissions to concerts and sporting events, etc.). The lawful gambling tax is not an additional or “sin” tax but rather is in lieu of the sales tax. The sales tax does not work as a tax on gaming, because it would act like a turnover tax as winnings are re-bet (a hefty state take-out). The lawful gambling tax addresses that by taxing the amount bet less prizes paid. It is the state’s basic tax on this consumption and as such should go for regular general fund purposes. One can quibble about what the gambling tax rates should be to mirror a sales tax but shifting entertainment spending to lawful gambling is unlikely to yield much, if any, of an increase in state revenues. It is simply the way this form of entertainment consumption pays state tax rather than through the general sales tax.
• The 2013 legislature added $126.5 million of unrelated (corporate and tobacco tax) revenues to the account. This money did come from new taxes but they were proposed by the governor, House, and Senate as general revenues. It was only at the 11^th hour of the 2013 session that they were designated for the stadium reserve to provide cover heading into the 2014 election.

Bottom line: the reserve statute wildly overcounts whatever the incremental general fund revenues resulted from enacting the stadium bill. It counts growth in revenues stimulated by inflation and economic growth as attributable to e-games, while failing to adjust for the reality that spending on e-games reduces other taxable spending. Authorizing e-games did generate some unknown general fund revenue growth, largely by diverting spending from tribal gaming to lawful e-games. The reserve’s calculations just make no effort to measure that amount, but rather capture a much higher amount.

#2: The account’s annual, rather than cumulative, calculations and the failure to adjust for the Minneapolis sales tax coming online allocate too much to the reserve.

The account determines whether to add money based on each year’s revenue and spending, but has no effective mechanism for making up for past deficiencies. This results in a “heads-I-win-tails-I-don’t-lose” situation (a ratchet effect). MMB’s releasing of money to cover 2015 and 2016 bond payments helps mitigate that effect but only partially. (As an aside, MMB had to stretch the statutory language to do that. The language authorizes prepaying or refunding bonds, but MMB used it to offset the cost of regular principal and interest payments; OLA surprisingly was not troubled by that in its audit of the account.) Its spending authority, for example, cannot be used to pay for ancillary costs of the stadium bill (e.g., appropriations for problem gambling programs or for grants to St. Paul). As its balance grows, there is no authority to recover the 2013 legislature’s allocation of tobacco and corporate tax revenues to the account.

The financial demands of the stadium bill on the general fund will drop dramatically when the Minneapolis sales tax starts to pay stadium costs (about $22 million per year). Because the account’s calculations consider the sales taxes as “available revenue,” that ensures the account’s balance will balloon, even if lawful gambling revenues rise only modestly. The state general fund has been paying the full cost of the stadium bill – loaning or advancing the Minneapolis share of the cost until its sales tax dedication to pay the convention center bonds ends. But the reserve account calculations do no change. That structure is a big factor in the jump in the account balance projected by MMB and reported by the media (see, e.g., Rochelle Olson, U.S. Bank Stadium reserve predicted to mushroom to nearly half a billion by ’25, Strib 12/3/20).

#3: The account is not limited to amounts needed for the bonds or stadium operations.

A standard bond reserve account would be limited to some fixed amount of future bond payments (e.g., those due in next one to two years). The stadium reserve has no such limit. Its growth is only limited by growth in lawful gambling and Minneapolis sales tax revenues. Why this is so is an artifact of its origins and the unique nature of appropriation bonds.

Bond counsel, the high priests of drafting bonding bills, insisted the reserve be divorced as much as possible from the bonds. That was to ensure the bonds did not look like debt, as defined by the constitution (obligations paid from a tax of statewide application), but rather as true appropriation bonds (un-debt under the constitution) whose only security and payment method was the bare appropriation. That prevented structuring the reserve like a typical bond reserve fund that would dedicate a modest amount for debt service payments due under a formula.

Of course, an absolute upper limit (like the governor’s proposal of a $100 million cap) could have been added. Had the possibility been raised, it likely would have been included in 2012 bill. But in 2012, none of the bill’s architects thought that accumulating 9-figure amounts was even a remote possibility. The consensus view was that revenues would fall short, as they did. (I know that as a drafter of the bill I gave it no thought and, thus, did not raise it as any concern with the authors. The account was such a minor feature no one spent much time or effort on it.) Of course, that was in the short run. Given the definition of available revenues (all growth in lawful gambling taxes and the Minneapolis sales tax) and the long time periods involved (30 years or so), that possibility should have occurred to us. Just inflationary growth over long periods yields a lot of revenue, putting aside the effects of the Minneapolis sales tax. (As an aside, the lawful gambling tax uses an unindexed progressive rate structure, yielding a bracket-creep inflation dividend in revenues.) But too often thinking and discussion focuses on the near term. So the lack of a limit was little more than lack of foresight by the designers of the bill.

Bottom line

The account’s calculations allocate materially more than incremental revenue attributed to authorizing e-gaming. Its flawed calculations set aside money in good revenue years but does not correct for down years. When the Minneapolis sales tax revenues become available, allocations to the account are sure to grow by a lot. Because it is untethered to bond payments and stadium costs, the account can (and under current projections will) grow to large amounts.

Possible justifications

The reserve is flawed, but one can similarly criticize many state laws and financial mechanisms. Retention of the account could be justified if it serves some other useful purpose. Three possibilities include:

• Protecting the original 2012 stadium deal.
• Allowing early prepayment of the stadium bonds.
• Providing a de facto additional reserve for the general fund.

None of these rationales supports continuing the reserve or, at a minimum, adding money to it.

Protecting the original stadium “deal”

This argument has been advanced. I assume that it largely means that additional revenues yielded by the stadium bill need to be used to pay the bonds or for other stadium costs. But the account’s calculations simply do not reflect that reality. The 2013 legislature added over $44 million of unrelated revenue to the account for which there is no mechanism to restore to the unrestricted general fund. The annual calculations significantly overstate any incremental general fund revenues generated by allowing e-games. The Potemkin Façade that these arguments comprise was useful to pass the stadium bill – a political necessary to cover keeping the Vikings here because of the difficult politics of pro sports subsidies. But now that the stadium is built and an established fixture, there is no need to continue the charade. Doing so sets aside more general fund money for the stadium than necessary. Why the Vikings would want to retain the reserve as constituted is obvious. Why the rest of us should is not.

Moreover, I am confident that if the 2012 legislature had understood the budgetary effects of the account that are now obvious, it would have modified or dropped the reserve from the bill. Keeping it in its current configuration does not preserve the original deal.

Prepaying the bonds

An often-advanced rationale for the reserve is to accumulate money to prepay the stadium bonds. That is one of its explicit statutory purposes. The bond issue allows redeeming bonds after June 1, 2023. Accumulating more in the reserve would provide a pot of money to do that, reducing future interest payments. This idea is superficially appealing but:

• Textbook public finance argues for spreading the cost of a capital project, like the Vikings stadium, over its useful life. That ensures that the cohort of taxpayers who benefit also pay. Prepaying the bonds will undercut generational fairness.
• Prepaying the bonds would save state interest payments, but the federal government subsidizes those costs, because the interest is exempt from federal income tax. Tax exempt bonds (most of the stadium bonds are tax exempt) are one of the least expensive ways to borrow money. Prepaying the bonds forgoes that federal subsidy, increasing how much of the stadium will be paid locally rather than nationally. The federal subsidy for borrowing is a strong incentive to spread the cost over time, helping match those who benefit with those who pay.
• Prepaying the bonds is not necessary to save interest. The call or redemption provision allows the state to realize savings by refunding the bonds if interest rates drop. Rates are now lower than when the bonds were issued in 2013. If that continues, interest savings could be realized by refinancing the bonds. The economic reality is that interest rates reflect the time value of money (discounted by the federal tax subsidy). Paying interest is not a real “economic” cost of the stadium. If the goal is to reduce interest payments, the same result could be achieved by using the reserve to pay cash for another new capital project that otherwise would be financed with bonds (especially if current interest rates were higher than on the stadium bonds – unlikely).
• I cannot recall the state ever just prepaying, rather than refinancing, a significant amount of bonds. That reflects the reality described in the three previous bullets.
• Prepaying the bonds will make it politically easier for the team to return to the legislature and ask for more, such as stadium upgrades. For those who think the state should minimize how much it pays in pro sports subsidies, that may not be a good outcome. As long as there are outstanding bonds, the public is more likely to resist calls for publicly financing improvements or enhancements to the stadium. Until the Metrodome bonds were paid off, calls for a new baseball park and/or football stadium were typically ignored by legislators.
• Prepayment could free up the Minneapolis sales tax for other city uses. (That does not necessary follow; a refinancing or prepayment could be restructured to keep the city on the hook for some or all its current share.) Depending upon one’s political perspective on how much it is appropriate for the city to pay, that may or may not be a good outcome.

Hennepin County funded much of Target Field’s cost by issuing bonds and is planning to prepay those bonds. Aren’t the considerations for the Vikings stadium bonds the same? The simple answer is that the two bond issues are different in an important way. Hennepin County imposed a new dedicated tax to finance Target Field – a countywide general sales tax. Prepaying Target Field’s bonds will allow that tax to die. The Vikings stadium bonds, by contrast, are paid with general fund money. No new tax was imposed. Authorizing e-games resulted in some unclear and ultimately difficult to determine additional revenue. That is why the account uses an easily measurable amount – the growth in all lawful gambling taxes over the 2012 forecast baseline. Repealing the account would simply reflect the reality that the stadium bonds are and always have been a responsibility of the general fund and that lawful gambling taxes are a general fund resource, just like revenues from the general sales tax, which they are imposed in lieu of.

A de facto general reserve

Squirrelling money away in the account could be considered a de facto extra general-purpose reserve since the legislature can always use it for other purposes when the budget is tight. The state currently has a large explicit reserve (over $2 billion). However, that it is not large enough to offset a major downturn. Two recent budget shortfalls (in 2002-03 and following the Great Recession) exceeded that by billions. Politically, it is simply not easy to maintain an explicit reserve big enough to protect against deep downturns. Having de facto or hidden reserves, like the stadium reserve, is useful budget protection.

I find this argument to be the best rationale for continuing to add money to the account. But it undercuts budget transparency and its designation as a “stadium” reserve puts a prior claim (implied entitlement) to stadium uses of its money, undermining its utility as a general-purpose reserve. For example, it is easy to imagine emptying the account to prepay stadium bonds right before a deep recession hits when more robust reserves are most needed.

Bottom line: None of these alternative justifications for the stadium reserve are even remotely compelling.

Conclusion

The account sequesters general fund revenues for presumptive stadium purposes for no good reason. It is not necessary legally or financially to support the bonds. It does not collect a dedicated revenue source or come even close to accurately measuring whatever incremental revenues were generated by the original stadium deal. It is time to move on.

Repealing the reserve (best approach) or capping it (governor’s proposal) would immediately free up money. It is worth noting that this includes ongoing or permanent revenue, not just a one-time amount, because the account claims all growth in lawful gambling tax revenue. Thus, the money could be used for ongoing purposes – whether spending or a tax cut, such as providing more federal conformity than the current global budget deal contemplates.

The account served useful political cover for the stadium deal – helping to make it look like general fund resources were not being diverted. Of course, they were. Since the Vikings are now unlikely to leave any time soon (they spent more than half a billion for their share of the stadium cost just a few years ago), everyone can safely move on. Compounding the original duplicity by allocating more money to the reserve is foolish.

Postscript – “fix” the account?

An alternative to repeal would be to “fix” the account – i.e., to measure the cumulative amount of incremental general fund revenue from authorizing e-gaming, to deduct the cumulative stadium bill spending, and to put the balance (if any) in the reserve. Authorizing e-games did provide growth in general fund revenues, maybe enough to pay for the bonds and other stadium bill spending (count me skeptical).

That would codify the political deal on which the stadium was sold. But it can be easily dismissed. Accurately measuring the additional general fund revenues from e-games would be exceedingly difficult to do (econometrically accounting for substitution effects etc.), of questionable validity, and controversial. The effort is not worth the candle. Moreover, even if we could magically know the right number for incremental revenues, there is no policy or principled basis for financing the stadium by expanding gambling. It was simply a political convenience.

Elephant in the room. I would be remiss if I did not briefly mention tribal gaming and its relationship to the stadium funding and the reserve. E-gaming clearly yields some new state tax revenue. Shifting spending from taxable entertainment (concert and sports tickets, bars, restaurants, etc.) to e-games generates little or no new state tax revenue but shifting spending from tribal gaming does. Federal law and the state’s agreements with the tribes provide that revenue belongs to tribal governments. Unlike some states, Minnesota’s agreements with the tribes do not provide for sharing revenues with the state.

If the political deal that funded the stadium did generate meaningful new state tax revenues, most of them likely came at the expense of the beneficiaries of tribal gaming – tribal governments and their members, as well as tribal casino employees and suppliers. Continuing the reserve and/or asserting stadium funding still depends on e-gaming revenues could create political/legislative issues related to regulating gambling and/or relations with tribal governments. I think that would be unfortunate:

• Tribes have special status as governments somewhere between a state and local governments, which are almost totally subordinate to the state.
• They are responsible for providing services to communities that are some of the poorest, if not the poorest, in the state. To help do that, federal law and the state agreements have given them a quasi-monopoly over casino gaming. But that additional revenue is still inadequate to address their needs and it is largely all they have beside federal aid. Yes, a couple tribes are well provided for (members of the Shakopee tribe, for example, are purported to have annual incomes for per capita payments derived from casino income like pro athletes) but they are the exception.
• Tribal communities derive no special benefit from the stadium. Why should they be expected to pay for much of its cost? A similar approach would be to single out a few rural local governments to pay – something no one would think appropriate. The fact that Indian communities heavily vote DFL likely clouds GOP legislators’ thinking on this. They fall into the trap of focusing only on the revenue side (casino profits) and ignoring the needs of tribal communities (poverty, poor infrastructure, etc.) – the same basis they use for regularly proposing draconic cuts in the state aid paid to center city governments (also heavily DFL supporters), which have generous tax bases but also large needs.

Decisions on regulating lawful gambling should be made on their merits, not based on some unknown effect on general fund revenues, much less a decade-ago decision to build a pro football stadium. The stadium connection does not appear to have entered the ongoing debate on rewriting e-gaming rules so far, but I can easily imagine it being raised and becoming an issue.

To understand how and why the state ended up with the peculiar and nonsensical (from a policy and finance perspective) reserve account, one needs to appreciate the political dynamics of state provision of pro sports facilities subsidies. These dynamics help explain how the Vikings stadium financing package was constructed, including the reserve. That history is in the first post.

DISCLOSURE: These are purely my amateur political observations. I was a longtime nonpartisan legislative employee and have never been involved in campaigns or even made a political contribution. So, I have no direct experience with the campaign dynamics that are involved. But as a legislative staffer I observed how legislators behaved and listened to many or their and partisan staffers’ private discussions on the politics of various provisions – it is what they often talk about in weighing policy options. For anyone who has been around the legislature and witnessed sports facilities debates, my observations in this post will be all pretty obvious, old news.

Side note: this post is purely about the politics, not public policy merits, of subsidizing pro sports. Much of the public debate (especially in the more distant past) consisted of extensive and heated debates about whether pro sports franchises generated incremental economic activity that could offset the cost of proposed government subsidies. Some of this occurred in debates during the Vikings stadium deliberations (mainly courtesy of Ted Mondale, Dayton’s point person in the discussions). There is a consensus among neutral economists that there is no such effect (see this now oldish article in the Journal of Economic Perspectives for a summary of that consensus). The main policy justification for the subsidies is for state and local governments to buy access to an amenity for their residents. I think most informed legislators recognized and accepted that reality during the Vikings stadium discussions.

The political reality of state and/or local government subsidies for sports facilities is a classic “No win” for elected officials or more accurately “You can only lose” proposition. It is driven by six realities:

1. Public subsidies are necessary for most metro areas (certainly a modest-sized one like the Twin Cities) to attract and retain pro sports franchises. I will not try to explain why that is so economically, beyond that the pro sports leagues are quasi-monopolies and their business models involve using the resulting power and the strong public interest in sports to extract government subsidies. This is done by insisting the public (not the users of the facilities, such as ticket buyers or the franchise’s service providers) pay for part of their private business facilities – arenas, ballparks, and stadiums – or they will move to a more accommodating metro area. Opponents of subsidies often contest this reality, but it is an axiom in varying degrees depending upon the pro league involved and the size of the metro area. The fact that the Lakers and North Stars left the Twin Cities is a local illustration of how sports leagues implement and enforce this reality. Moreover, most elected officials are unwilling to risk calling the leagues’ bluffs, even if they think the axiom is not true in a specific case. As a result, revenues that are anything close to facility-based or team-related are off limits – no tapping naming rights, ticket taxes, facility parking revenues, suite sales, or similar to pay for the public share of the facility cost. Related revenues from other businesses, like taxes imposed on the hospitality industry in close proximity to the facility, might be okay but that is about it and usually that is far from enough to generate the money needed – especially for an MLB ballpark or NFL stadium in a market like the Twin Cities.
2. The public – but especially a core group of important swing voters – wants and thinks the state or metro area is entitled to pro sports teams. It is a given in most folks’ minds that the Twin Cities and Minnesota should have an NFL, MLB, NBA, and NHL franchises. It is an expected amenity. Support or enthusiasm for the local teams, especially when they win or have national reputations, is a big deal and helps bind the community together – often across increasingly polarized partisan lines (one of the few areas of society where that still occurs). The support is particularly strong among working-class, white males. They are a coveted constituency for both parties; substantial majorities of them swung from Reagan to Obama and back to Trump. Most elected officials in both parties – particularly those who are more centrist and/or who understand the key political dynamic of attracting the “median voter” to gain or stay in power – recognize this reality. Also, elites (e.g., corporate executives) expect that the area should have pro sports franchises, because it makes the area more attractive for their employees and customers (regardless of how they personally feel about pro sports). That combination (working class males and elites/donors) makes finding a way to keep the pro sports franchises very important to key elected officials in both parties, regardless of whether they think it is an appropriate function of government.
3. Core activists in both parties strongly oppose paying those subsidies. While most elected leaders of both parties recognize the financial, business, and political reality of the need to pay the subsidies and that large groups of key swing votes want the teams to stay put, most activists in both parties passionately believe philosophically that the subsidies should not be paid. Right wing activists (Republicans) tend to have libertarian-oriented views and think that pro sports, like any other business activity, should be a private market activity. Left wingers (Democrats), by contrast, are fine with big government but strongly believe that providing bread and circuses is not an appropriate government function (particularly when they subsidize billionaire owners and millionaire players – subsidizing starving artists might be okay). So, there is a bipartisan consensus of sorts opposing the subsidies. These folks work on both parties’ campaigns, pay attention to the details of legislation, write letters, attend meetings, care passionately, and so on. They are the infantry of both parties’ armies. Thus, elected officials strongly want to keep them happy – if only to avoid primary challenges and questioning of their ideological bona fides. Elected officials on the ideological fringes of both parties share their opposition and are more than happy to stoke conflict in the legislature and public debate generally.
4. The public agrees with them. Faced with the question in the abstract – Should state government subsidize pro sports? – the public agrees. This is verified by polling (as I recall without bothering to dig up specific polls) and creates a disconnect with points ## 1 and 2.
5. The media focuses a lot of attention of the issue. Pro sports is a big business and a matter of public interest – witness the news hole the typical major newspaper devotes to sports coverage, as well as extensive network and cable television coverage. The public subsidy issue is easy to cover; it is simple to convey that government is handing lots of money over to rich private business owners by building shiny new facilities for their teams. The media extensively covers the issue with many stories. That generates interest and knowledge, stoking public opposition (for the reasons noted in ##3 and 4) and makes it difficult to quietly slip a financing package through.
6. If a team leaves, elected officials will be held implicitly responsible. This seems unfair but is the hard reality that governors and legislative leaders recognize. Making matters worse, if a team leaves, it is widely accepted that attracting a new one will cost even more in government subsidies and will take a long time.

Bottom line: It is a no-win proposition for elected officials in both parties, but mainly the ones who are in control (governor, speaker, and majority leader). The public and their own party activists do not want them to provide subsidies. But it is also fair to say a healthy majority of the public, especially white male swing voters, want and expect to have pro sports franchises. Put even more starkly, a sizable and key portion of the electorate feels entitled to local pro sports franchises, is unwilling to have their taxes pay for the privilege, but will hold elective officials responsible if a team leaves (especially the Vikings, one would guess since NFL football is the most popular pro sport by a wide margin). Moreover, business elites (key contributors and opinion makers) strongly think legislators and other elected officials should be able to navigate this mine field and preserve what they also consider a basic metropolitan and state amenity.

This reality is clearly revealed by the stark contrast between the 2011 and 2012 legislative sessions. The 2011 legislature and governor engaged in a protracted and bitter partisan budget battle that resulted in a government shutdown over tax increases versus spending cuts. Without any material change in the budget situation, the same legislature and governor in 2012 reached a bipartisan deal to spend a half a billion in tax revenue on a Vikings stadium.

The net effect of these political dynamics is that:

• Leaders in both parties have a big incentive to keep the teams satisfied, if not happy. That is why they are so determined to get deals done (or allow them to be done) despite opposition by many of their most loyal and activist supporters in both parties and, perhaps, their personal skepticism about the wisdom of doing so.
• Financing should be low profile to avoid as much public attention as possible and to make it look like general taxpayers are not really paying – hence, politicians will falsely say users are being tapped or new or extra revenues that (magically) do not come from general revenue sources (taxpayers) are paying.
• It is easiest to do that when split government enables a sort of bi-partisan, nonaggression pact (at least at higher levels). When only one party is in control (and this is crucial in the reserve account history), the other party will feel free to challenge, oppose, and criticize their actions. Either party can do so consistent with its philosophy and principles and will be cheered on by their activists and much of public. Doing that is much harder to do when the actions have broad bipartisan backing.

A few specifics on identifying revenues. The core political challenge is to find a source of revenue – that is both necessary to pay the bonds and is also the way voters processes who will pay. The public (based on most polling) will accept having users and beneficiaries pay. But the whole premise of public involvement is to get taxpayers, not the teams, to pay. That puts legislators (and the governor) in a pickle – how to come up with the money to build the necessary facilities without appearing to make the general public pay. In short, they are on a hunt for Rumpelstiltskin.

The typical approach is to turn to one or more of the following options:

1. Passing the political buck down to a compliant local government, which uses a reliable local tax source – that is what was done to finance the Twins stadium with the Hennepin County board (a powerful but politically low-profile body) taking the heat for imposing a county sales tax to finance Target Field.
2. Taxing (mainly) tourists and other nonresidents through taxes on rental cars, hotels, and so forth. That has been a popular approach in other states, not as much in Minnesota because the state has already tapped those sources to the max for other purposes.
3. Taxing sin – booze, cigarettes, and/or gambling. This proves harder to do than a casual observer would expect. Purveyors and enjoyers of vice don’t like to pay taxes any more than the righteous and they vote (purveyors have lobbyists too). The Metrodome was largely financed with a local liquor tax, which was highly controversial and narrowed almost immediately after it took effect. Former Governor Arne Carlson unsuccessfully tried to finance a new baseball park by hiking the cigarette tax and Republican legislators repeatedly tried unsuccessfully to use a racino (authorizing and taxing casino gambling at the racetracks) to finance a variety of facilities and purposes, including sports facilities.
4. Using “new” state tax revenues justified by an assertion that some newly authorized activity is paying, holding taxpayers and beneficiaries of government services harmless. Assertions that the revenues are “new” or “incremental” may be questionable or outright wrong, but if enough people (e.g., key legislators and news media coverage) accept the assertion, it does not matter.

The Vikings stadium deal relied on a combination of ##1, 3 and 4 – capturing the future revenues from a local Minneapolis sales tax and authorizing a new form of lawful gambling (e-games) that Rumpelstiltskin would spin into gold. The main message was no new taxes would be imposed and no existing general fund revenues diverted. The latter was not true as an economic matter. A large but unclear amount of money that otherwise would have gone for general fund purposes (or enabled cutting taxes) was diverted to pay for the stadium and related costs. (I cover why I think that is so in more detail in my third post.) But the deal worked in that it solved the political problem of building the stadium and keeping the Vikings here without any apparent effect on the 2012 or 2014 legislative or gubernatorial elections (at least as far as a casual campaign and election observer like me could tell).

The reserve account plays into this because the Dayton administration’s (and DFL legislators’ who approved them) 2013 changes to and subsequent management of the account were made to bolster that political narrative. That is so because they added $100+ million in new money to the account’s calculations and treated the account as essentially a running scorecard validating that the deal generated enough new revenue to cover the stadium’s public costs. The positive balances in the account presumably reflected that the deal was generating new money (mainly from e-games) to cover the costs.

As an aside, using the account in that way may have been their plan already in 2012 (i.e., not just because of the 2013 changes). That was certainly inconsistent with my understanding of intent of the original account and bond counsel’s explicit directions in 2012 to make sure that the account would not be view as linked to, paying, or securing the bonds. The intent doesn’t really matter now. What does is how the account is now being treated as the ad hoc bond payment fund.

While this may have worked politically (the media continues to buy it as well as the public, I assume), it suffers from two major flaws as described in my third post:

• It is factually wrong; the account’s calculation do not measure new or incremental revenues that “but for” the stadium bill would not have been collected by the general fund. Not surprisingly, the thumb-on-the-scale calculations used for the account materially overcount those revenues.
• The account, as a result, sequesters money for the stadium that is not needed (nor was it part of the original “deal”) and that distorts general fund budgeting.

Since the controversy over whether to build and how to fund the stadium is now a distant memory and the key players (governor, speaker, and majority leader) have all left the scene, there is little need to continue this political charade. (Yes, the senate author, Senator Rosen, and Senator Bakk, an undoubted player in the 2013 changes, are both still in office. But the details of an almost decade-old stadium deal seems unlikely to factor into whatever future political plans they may have – probably nothing beyond retaining their legislative seats, if even that.) Axing the account will generate a few media stories that everyone will quickly forget. Repeal would certainly not become an election issue. It is for these reasons that I probably naively think that political considerations should no longer be an impediment to repealing the account.

This post recounts the history of the 2012 stadium deal, including the reserve provisions and the 2013 changes to them, a history that helps understand the reserve, which can only be described as unusual. The second and third posts delve further in the politics and policy involved.

DISCLOSURE: I was one of the drafters of the 2012 stadium legislation. As a result, I am not an unbiased source. My views are clouded by the perceptions I developed during that process and working on other pro sports facility finance bills going back to the bill that financed the Metrodome in my first legislative session (I had a minor role working on its bonding provisions).

In December, I was certain that budget exigencies would drive the 2021 legislature to revise the reserve. A sizable deficit with the no-tax increase Republicans controlling the Senate would make the account’s growing pot of money a tempting target. Governor Walz’s proposal to tap the fund (via a $100 million cap) reinforced my view. The February forecast of a surplus and the large aid fed aid infusion now make that less likely. The House omnibus tax bill includes the governor’s recommendation, so action remains possible. But if nothing is done in 2021, the issue will not go away.

The stadium reserve is not your father’s bond reserve fund (aside from the fact that it is an “account” not a fund). A typical bond reserve protects bondholders and ensures the bond’s obligations are paid on a timely basis by setting aside (typically dedicated) revenues for the project and its bonds. The stadium reserve is none of that because of the unusual nature of its bonds. It is a purely political construct whose financing morphed within a year of its creation. It was neither legally nor financially necessary to protect the bondholders or to construct the stadium.  Whether it should continue in its current configuration or at all is, thus, a political and policy question.

The post consists of three parts:

1. A description of how the stadium financing was constructed and works. This is essential to understanding why the reserve is legally and financially irrelevant to payment of and security for the stadium bonds.
2. An account of how and why the account was included in the 2012 bill and changed in 2013. Contrary to its inherent nature, MMB has used the account and the media and public have viewed it as integral to the stadium financing – essentially as the account that pays for the stadium.
3. A list of the account’s effects.

Part One: 2012 stadium financing

Pro sports facility financing basics

Successful pro sports facility financing legislation needs to work on three levels. All are obvious but it is still useful to keep in mind that a proposal must:

• Identify and set aside enough revenues (taxes, fees, or charges) to pay for the bonds (debt) that financed its construction. Theoretically the government could pay cash to construct a major pro sports facility, but that is never done in my experience. It has to work financially.
• Be a constitutionally permitted use of the government’s money. It has to work legally. If bonds are used, as is almost always the case, bond counsel must be willing to issue an unqualified opinion to that effect. Bond counsel are independent lawyers, a private law firm on a list of firms that Wall Street accepts as having sufficient professional competence and rectitude that their unqualified opinion all but ensures the bonds’ legal status. That means the state’s designated bond counsel is essentially the high priest who determines the legality of the bonds (or the voodoo doctor who reads the chicken entrails that are the state constitutional provisions on bonding). You may disagree or think they are being too conservative, but if you cannot change their minds, it does not matter.
• Attract the necessary votes in both houses of the legislature and the governor’s signature. It has to work politically. To do that it has to be acceptable to the public, as perceived by legislators. The second post describes my views of the politics of financing pro sports facilities. Political considerations are a (probably the) key driver of the financing structure. That is, of course, the case for all controversial legislation.

Plans typically have two structural components, ignoring the unlikely possibility that the government will pay cash:

• Revenues – a stream taxes or fees
• Borrowing authority – the legal ability to issue bonds that investors will buy

So, architects of stadium financings must devise a plan with both of those structural elements, each of which checks all three functional boxes. Political acceptability is typically the most difficult nut to crack. Plans can easily identify sufficient revenues (even for billion-dollar stadiums), but opposition to tax increases and spending money for billionaire owners and millionaire players on both the left and right makes doing so a challenge. It is not as difficult as solving Rubik’s Cube blindfolded, but close. The large price tags for NFL and MLB stadiums and intense media interest in pro sports make them very visible politically and particularly difficult to do; sports arenas (like the Xcel and Target Center) are difficult but more doable because of their lower cost and visibility.

The key challenge is to identify a politically acceptable stream of revenues. Those revenues are the way the public (voters) process who will pay. They generally will accept having users and beneficiaries pay. But the premise of public involvement, as outlined in the second post, is to get taxpayers, not the teams and ticket buyers, to pay. That puts ticket taxes, naming rights, team rent, and most facility-related revenue off limits and legislators in a pickle – how to come up with the money to build the necessary facilities without appearing responsible for imposing taxes or cutting other public spending to do so. Because user fees do not work (recall the sine qua non of a plan is come up with new, non-facility related revenues) and clear or obvious state tax increases do not work (they are politically unacceptable), the typical approach is to turn to one or more of the following:

1. Passing the political buck down to a compliant local government, which uses a reliable local tax.
2. Taxing tourists and other nonresidents through taxes on rental cars, hotels, and so forth.
3. Taxing sin – booze, cigarettes, and/or gambling.
4. Using “new” state tax revenues justified by an assertion that some newly authorized activity is paying, holding taxpayers and beneficiaries of government services harmless.

The legislature turned to a combination of 1, 3, and 4 to finance US Bank Stadium. But the nature of those revenues as they intersected with legal restrictions on state bonding created complications.

Outline of the 2012 stadium financing

Stadium finance legislation always involves debates about where to construct and how to configure the stadium, interjecting issues of local politics, design preferences, and a host of other considerations. These issues greatly complicate the challenge, but I will skip over them as irrelevant to the financing structure and reserve account beyond noting the decision was made to keep the stadium in downtown Minneapolis because that was where it was, it avoided the need to acquire a new site with all the complications that entails, and it conveniently provided a possible revenue source. A book could easily be written about all the options and back-and-forth that occurred over these issues.

The legislation used two revenue sources, each of which had serious drawbacks:

• Minneapolis 0.5% general sales tax – this tax had been imposed since 1986 to pay for the Minneapolis Convention Center. Thus, it was a reliable source of revenue and falls into category #1 above, allowing passing the political buck down to the city. The complications were twofold: (1) it was pledged to pay convention center bonds through 2020, so it would not be available for years to pay debt service and (2) support for using it by city officials was uncertain. City government is very visible politically, unlike county government, and Minneapolis voters were opposed; at least they had amended the city charter to prevent such uses years back, a limitation that state law could nullify. But council approval was required by the state constitution, as implemented by a preexisting general law, and getting was no slam dunk (to use a sports metaphor).
• State lawful gambling taxes – the idea was to allow charities that run lawful gambling operations to offer electronic games and to capture the increased state tax revenues. These revenues straddle categories ##3 and 4 above. The organizations had long wanted authority to offer e-games but had been stymied by anti-gambling and tribal gaming interests. The lure of a Vikings stadium could overcome that arguably. However, two big difficulties loomed: (1) uncertainty as to the yield because there was little experience elsewhere with similar electronic games and (2) constitutional rules would require state bonds paid with states taxes to pass the legislature by supermajority (60%) votes, moving the political difficulty dial to the impossible mark.

To recap, these revenues had three main limitations:

1. A timing gap for the Minneapolis sales tax – its revenues would not become available until 2021 when the convention center bonds were paid off, half a decade after the stadium was to be constructed. Equally important its revenues were insufficient to finance the public share of the Taj Mahal the Vikings wanted, even had they been available immediately.
2. Uncertainty and likely inadequacy of the state tax revenues from electronic gambling – given the lack of a track record, conservative municipal bond investors were unlikely to buy bonds backed by them.
3. The constitutional supermajority legislative vote requirement to approve state bonds backed by or paid with a statewide tax, such as the lawful gambling tax.

Each of these barriers looked fatal, but luck and determination to find a solution by key legislators and the governor yielded a solution. Luck came in two forms:

• The first was a providential fiscal note on the effects of allowing electronic games prepared by the staff of the Gambling Control Board. It concluded that the games would be immensely and immediately popular, yielding a billion-dollar increase in lawful gambling gross receipts. (Its conclusions were largely based on projections made by a firm that was peddling the electronic gambling equipment.) Legislators and legislative staff (as well as others) were skeptical, but DOR Research, the agency responsible for scoring tax changes, reluctantly agreed to produce an estimate of the tax revenues based on the fiscal note’s assumptions about the popularity of the games. The revenue estimate dryly noted that “assumptions were provided by the Gambling Control Board[.]” (p. 5). This piece of luck “solved” the problems of revenue inadequacy and the timing gap resulting from the unavailability of the Minneapolis sales tax revenue until the convention center bonds were paid. Solved is in scare quotes because the solution was a technical one – i.e., fiscal notes and revenue estimates are legislative scores that enable passage of the law, but bond buyers were never going to rely on such flimsy projections. Solving that and the supermajority approval problem depended on the second piece of luck.
• The 2011 legislative session featured a multi-billion-dollar budget deficit that resulted in a political deadlock and government shutdown. The GOP-controlled legislature refused to agree to the governor’s tax increases, while DFL Governor Dayton refused to agree to their spending cuts. The natural resolution: deficit spending. It is a truism that Minnesota, like almost all states, cannot run deficits – that is, it cannot borrow money to fund ongoing operations beyond what is necessarily to smooth out biennial cash flow needs. The legal details of that truism rest on the constitution’s limits on debt. Very few constitutional limits, of course, are absolute and government shutdowns and implacable political deadlocks call for desperate measures. In 2011, the one devised was to use a technique that avoided being debt under the constitution but that was for all practical purposes – that is, appropriation bonds. Whether that worked was unclear, but bond counsel was convinced it would contingent on a Minnesota Supreme Court test case that was in process. Because the nature and characteristics of appropriation bonds are at the core of reserve account issues, more detail is provided about how they work.

Appropriation bonds are like a chameleon. For constitutional purposes, they are not debt, while for practical and financial purposes they are. The security for payment of appropriations bonds is a statutory appropriation (see section 16A.965, subdivision 8, for the appropriation for the stadium bonds). Under the constitution (article XI § 1), money can be paid out of the state treasury only if an appropriation permits it.  A line of court cases holds that lack of an appropriation (including repeal of one) prevents the state from paying an otherwise legal obligation (e.g., for services rendered to the government under a contract). This combination allowed issuing bonds backed by only an appropriation and disclaiming the bonds are a state debt obligation (see section 16A.965, subdivision 6), avoiding the constitutional limits on debt – in the view of bond counsel IF the supreme court agreed. (The court validated using appropriation bonds later that year. Schowalter v. State, 822 N.W.2d 292.) That followed from the technical ability of the state to renege by repealing the appropriation. BUT that is not a realistic option since it would cause a serious stain on the creditworthiness of other Minnesota state bonds and likely lock the state out of future borrowings for any purpose or at very high interest rates. Thus, the state has little practical alternative to paying the bonds. The bond market’s perception of this impracticality causes the interest rates on the bonds to be only slightly higher than the rates on traditional state general obligation bonds. The even better news is the bonds could be authorized by a simple majority approval, not the supermajority (3/5) required for most state general obligation bonds.

The legal basis for paying these bonds, thus, is a naked general fund appropriation, which a future legislature could repeal. There is no pledge of state tax revenues (that would make them state debt) or a mortgage on property (e.g., on the publicly owned stadium). The law was drafted, at the direction of the state’s bond counsel, to be clear on those points. That avoided any appearance of setting aside or reserving state tax revenues that could contaminate their status as appropriation bonds. That was not true of local taxes, such as the Minneapolis sales tax, because the constitution refers to “taxes of statewide application” in defining debt. Pledging local taxes would be okay. But the bonds are paid out of and secured by a general fund appropriation as a matter of law and that appropriation is their only “security” not a stream of tax revenues. Thus, the reserve needed to be structured to avoid appearing to pledge, dedicate, or use state tax revenues as the bonds’ payment mechanism.

Resolving the political problems. Appropriations bonds were a solution to the legal and financial problems of financing the stadium. But potential political problems remained. All the important elected officials had been clear that the stadium was not to be paid by the general fund and that, of course, is what appropriation bonds rely upon. That is a problem only if sufficiently important politicians and/or the media make it an issue. The solution was to construct a Potemkin Village. The Potemkin Village’s facade was the assertion in the providential fiscal note and revenue estimate – that is, that the increased lawful gambling taxes from e-games would be so robust that uses of preexisting general fund revenues would be unaffected. That assertion was questionable. Whether sufficient usage of e-games would occur, in the short run, or how much the resultingly revenues were truly a net gain to the general fund, in the long run, was unknowable and seemed unlikely to be enough. (Those two points are discussed in my third post.) But so long as everyone of any importance agreed to only look at the façade, it would work politically. The consensus on the need to build a Vikings stadium to keep the team and the recognition that that required a bit of financial magic or legerdemain were undoubtedly big factors.

As an aside, recall that this occurred as the state was struggling to recover from the Great Recession. The previous session has seen a protracted, partisan dispute over the budget resulting in a lengthy government shutdown. The fact that the two parties could come a stadium reveals how important they considered resolving this to be.

Key features

These are the key features of 2012 stadium deal (rough annual numbers are in a table after the bullets)

• Issue state appropriation bonds to pay for the public share (about $500 million); team finances the rest of the billion-dollar cost (ultimately it upped its share to add features to the building). A key sticking point in the negotiations – until the very eleventh hour – was how much the team would pay.
• Capture the Minneapolis local sales tax after the convention center bonds are retired in 2020 – the state collects this tax and would retain the revenues to  pay the city’s share of the state bonds (about 30%), plus stadium operating costs and contributions to the capital reserve. The Minneapolis tax pays more than 40% of the public cost – bonds and operating costs combined – once it becomes available.
• Authorize lawful gambling organizations to offer e-games and restructure the lawful gambling taxes (shifting who pays without increasing state revenues beyond the effect of e-games); the increased general fund tax revenues were assumed based on the providential fiscal note and revenue estimate to cover the amounts in the table. When the Minneapolis sales tax came online it would generate over $20 million per year, easing the demands on gambling tax revenues.

Spending item	Amount (millions/yr)
Total debt service	$30
State share	21
Mpls share	9
Operating and capex (Mpls)	7.5 (indexed)
St. Paul grants & problem gambling (state)	3.4
Total	$40.9

Part Two: The Reserve Account

2012 version of account

Since appropriation bonds were used, there was no legal or financial rationale for a reserve account. The bonds would be paid with a general fund appropriation, a draw on a multi-billion-dollar fund with vast revenues. Bond counsel wanted the law to be clear that the appropriation was the only basis for paying and securing the bonds. The language was to be bereft of any suggestion that state taxes were pledged, set aside, or used to pay the bonds. Obviously, most general fund moneys come from taxes, but because money is fungible the law could ignore that inconvenient truth. Once deposited in the general fund, there is no difference between tax revenue and nontax revenues such as fees. Wall Street also did not need any assurance that moneys were being set aside; it was relying on the reality that the state would pay or become a bond market pariah.

So, why did the reserve account get added to the law and what function did it serve? The 2012 legislation created the reserve account to address two contradictory issues:

• What should be done with all the tax revenues that were projected to flow from e-games? The providential fiscal note generated a gusher of hypothetical revenue – $58 million/year, much more than needed to pay for the bill’s spending (stadium bonds, its operations, ancillary grants, etc.) even before the Minneapolis sales tax revenues were available.  Key legislators and staff were skeptical that those revenues would materialize. The sensible thing would have been to ignore them (i.e., leaving them on the bottom line if they showed up), but legislatures rarely take sensible approaches. Rather, they follow their set processes and procedures – in this case, specifying how the revenues would be used. Lawful gambling interests insisted the revenues should be used to cut rates in the restructuring of their taxes. GOP legislators were happy to oblige but the administration sensibly refused to do so. (As noted above, a 2012-13 budget gap had been covered by borrowing. Potentially adding to that would have been the risky.) The compromise was to enact a contingent tax rate reduction, if revenues came in as estimated. But that was an all-or-nothing proposition – i.e., the revenues need to hit a specific mark to pay for the proposed cut in the tax rates. If revenues were less than that (or even more!) they still needed a specified purpose – viola, the reserve account whose uses were restricted to stadium purposes.
• How can we hold the Vikings partially responsible for possible shortfalls in funding? At the same time that they were debating what to do with the supposed surfeit of revenues from e-gaming, other legislators thought revenue shortfalls were more likely and wanted contingency plans to address that – especially to require the team to pay more. This was done by providing contingent authority to impose a tax on stadium suite rentals after a (meaningless) sports-theme lottery game was offered. (Everyone recognized adding a lottery game would not generate more revenues since the lottery was already maximizing its sales.) The reserve account calculations provided a convenient way to do that.

The reserve account addresses these two contradictory goals. It is implemented by an annual calculation of “available revenues” – a rough (and inaccurate on the high side) proxy for the increase in revenues generated under the bill’s provisions – less the expenditures the bill authorized. If this number was positive for a year, MMB sets that amount aside in the reserve. The account was in the general fund and was not required to be a real account (e.g., a separate bank account or similar) but can be just an ongoing annual reconciliation of the amounts.  That is how MMB administers it – essentially as a spreadsheet of notional amounts. See Minn. Stat. § 297E.021 (2012) for the original language.

The statute explicitly says it is not a “pledge” of the available revenues to pay the appropriation bonds. See subdivision 1. The statute (subdivision 4) appropriates its money for “application against shortfalls” in the Minneapolis sales tax (using local taxes to pay the bonds does not trigger the constitutional definition of debt because they are not “statewide” taxes) and “for other uses related to the stadium” as the commissioner determines prudent. The only specifically authorized spending related to the bonds is to pay “refundings, and prepayment” of them. By negative implication, it is not for regular principal and interest payments.

So how do these provisions help to implement the contingent reduction in the lawful gambling tax rates or to hold the Vikings responsible for revenue shortfalls?

• The contingent reduction in the lawful gambling tax rates was keyed to revenues in fiscal year 2016, as provided in section 297E.02, subdivision 6(b) (2012). The amount triggering the reduction was set based on the 2012 revenue estimate and projected stadium costs (i.e., the list in the reserve account statute). If sufficient collections had occurred, it would have automatically reduced the calculated “available revenue” under the reserve account statute. Of course, the revenues did not increase sufficiently, and 2016 passed without implementing the reduced rates. The Revisor removed the obsolete tax cut language from the statute.
• The authority to trigger a special suites tax is linked to a truncated portion the reserve account calculations (growth in lawful gambling tax revenues compared to direct stadium costs – not problem gambling and St. Paul grants). If there is a shortfall, the statute gives the commissioner of MMB discretion to impose a 10-percent tax on stadium suites after implementing a sports-themed lottery game (the consensus again was that the lottery game would not raise new revenue).

In conclusion, the addition of the reserve account to the 2012 legislation was a side effect of the unreal revenue estimate and a thrown-in deal point that gave the appearance of holding the Vikings’ feet to the fire if the projected lawful gambling tax revenue did not appear. During the 2012 legislation session with a Republican legislature and DFL governor, agreement on the financing plan (aka the façade) was done on a bipartisan basis. At that point, the reserve account was not much of a factor; it was created for other reasons. A side effect (I was never sure if was intended or unintended) is that the account’s calculations provided an ad hoc (and highly inaccurate, as I explain in the third post) scorecard on whether e-gaming revenues were paying for the stadium, as promised. That gave supporters or opponents (depending upon how revenues flowed) something to point to in claiming either the deal had worked or failed.

2013 changes

The DFL swept the 2012 election, regaining control of both houses of the legislature for the 2013-14 legislative session. That was just in time for the cracks in the stadium financing’s Potemkin façade to appear; revenues from newly authorized electronic gambling were clearly not going to be sufficient to cover the cost in the early years. Lawful gambling revenues fell rather than rising by the estimated tens of millions. (FY2013 lawful revenues, the first year in which effects could start to show up, were lower than for FY2012. Projections were not any more optimistic – nothing like the over $40 million bump the stadium deal was premised on.) That did not create a legal or financial problem; payment of the yet-to-be-issued appropriation bonds did not depend on gambling tax revenues. Wall Street would not have been concerned. But it created a political problem because now the public could easily see around the Potemkin façade as the news media was sure to point out.

That political problem was no longer bipartisan but owned by the DFL since it controlled the legislature and governorship. More importantly, Governor Dayton was up for reelection in 2014. A Republican candidate for governor could easily charge him with flubbing the stadium deal in multiple ways – using general fund money which he had publicly pledged not to do, failing to make the Vikings and users pay enough, identifying an unworkable revenue source, expanding gambling, and so on. (As an aside, this might not be a problem if his Republican opponent were a legislator who had voted for the deal but that was unlikely and turned out to not be the case. Former Speaker Kurt Zellers was exploring and did run for the Republican nomination, but he had conveniently voted against the stadium deal. His tacit support as the speaker in 2012 was likely crucial to allowing the deal to go forward, but that sort of nuance is lost in a political campaign.) Because of the unusual politics of pro sports facility financing (the public believes it is entitled to keep the teams but does not need to pay for doing so), it was easy to see how the stadium could become a political problem for Governor Dayton in 2014 or for some DFL legislators who had voted for it and represented competitive districts.

To address these problems, the governor and DFL legislators turn to the reserve account for a political fix. The solution, as far as I could tell, came purely from the Dayton administration (the administration presented it to the conference committee, and I was unaware of legislative involvement in formulating it but DFL legislators were likely consulted at a minimum). It made two additions to “available revenues” under the reserve account statute:

1. A one-time addition of $26.5 million. This money came from imposition of a floor stocks tax on cigarette and tobacco products inventory that complemented the 2013 legislature’s increase in the ongoing excise taxes.
2. “Up to” $20 million of corporate franchise tax revenues each year. The statute does not say this, but the justification for the $20 million amount was a revenue estimate for a proposed change (tax increase) in how the corporate tax is calculated, specifically a modification of the apportionment formula for unitary groups with members that did not have Minnesota nexus. The $20 million is fixed and does not depend upon the accuracy of that revenue estimate (accuracy which would be very difficult to verify, in any case).

These changes were proposed and made at the eleventh hour of the legislative session – during the final days of the tax conference committee. The changes appear simple and minor; not surprisingly, they attracted little attention. The reality was otherwise.

Impact of 2013 changes:

1. Acknowledge the 2012 deal had not identified sufficient new revenues to pay for the stadium. The changes formally recognized that authorizing e-games would not generate enough revenue to pay for all the deal’s costs. This might be a short run problem until the games were implemented and became popular and/or until the Minneapolis sales tax became available. But it was a shortfall. How much and how long was unclear but the changes added at least $26.5 million one-time plus $20 million per year for an indefinite period. As it turns out, the amount was $126.5 million.
2. Identify “new” general fund revenues for stadium uses. These were new revenues because they derived from 2013 tax increases, but they had no connection to the stadium. Both taxes were included in the Governor’s budget and in the tax bills that both houses passed as general revenue raisers. The reserve account changes restricted their uses to stadium purposes, a restriction adopted at the very end of the legislative session.
3. Add administrative discretion. The original reserve account language provided straightforward accounting calculations – the defined “available revenues” less the listed expenditures. The 2013 language, by contrast, added to the available revenues “up to” $20 million of corporate tax revenues. The “up to” must have meant the MMB commissioner would decide how much was needed – something the statute itself does not say. Subsequent practice showed that the administration meant that the $20 million (or less) was to be added to the account’s listed revenues if the other revenues were less than the listed spending. But the statute’s language does not specifically say what “up to” relates to or is limited by.
4. Help the reserve become the de facto stadium funding account. The 2012 version of the account addressed the remote concern that a gusher of e-gaming tax revenues not go to pay other general fund spending. In 2013, the opposite problem occurred (not enough revenues) and the DFL wanted to modify the 2012 deal by coming up with more revenues so that the stadium would not be perceived to be diverting general fund resources. But doing that in the law was inconsistent with using appropriation bonds, because designating state tax revenues as paying them could jeopardize their “nondebt” constitutional status. They could have done that rhetorically but obviously wanted to put a marker in the law to point to in political arguments. So, they turned to the reserve account. Adding discretion and ambiguity to the language probably helped them to administratively run the account like a dedicated fund for the stadium. Because bond counsel needed to sign off (the bonds had not been issued and the state was still waiting for the Minnesota Supreme Court to say they were a valid “nondebt” financing technique), the language was cryptic and ambiguous, flexible enough to preserve the legality of the bonds while giving the administration political cover. All it did was change the financial computations. The whole ruse worked because the court decided favorably, the bonds were issued, and the stadium was built. The administration proceeded to treat the reserve account as the stadium funding account. Releasing money (that the 2013 changes had added) from it in years when lawful gambling revenues fell short. This practice is documented by the Office of Legislative Auditor’s financial audit of the reserve account, which reports that money was taken from the account to pay stadium costs in two fiscal years (2015 and 2016). The ambiguity in the language was sufficient for these practices to avoid sanction by OLA, even though they were inconsistent with a careful reading the original purposes for the account (in my view).

These changes provided a way to deflect objections to the stadium financing in the 2014 election campaign. The stadium had not been a big issue in the 2012 election, which occurred shortly after the bill passed. But now that the projected increase in gambling tax revenues had not appeared and preexisting general fund resources would be more clearly paying, objections and campaign charges seemed likely to be made. The plan worked. The court decided favorably, the bonds were issued, the stadium was constructed, and its financing did not become a significant 2014 campaign issue. Given the complexity and opaqueness of the mechanism, the media bought this narrative and have essentially treated that stadium as being paid with dedicated lawful gambling revenues augmented with some one-time other tax revenues.

Part Three: Its effects

2012 version. The account was largely a minor after thought in the 2012 stadium bill and was not essential to issuing the bonds or building the stadium. Authorizing appropriation bonds ensured that. Rather, it was intended to collect extra lawful gambling revenue if the providential fiscal note proved accurate but yielded less than enough to cut the tax rates as proposed in the bill. Since actual tax collection numbers through fiscal year 2020 are available, we know now that no money would have been set aside in the reserve under its original provisions until fiscal year 2019 when $11 million would have been designated for the account.

2013 provisions. The 2013 changes allocated $126.5 million more to the account – the $26.5 million of cigarette floor stocks tax revenues and $20 million/year of corporate tax revenues for fiscal years 2014-2018. MMB’s use of the reserve to pay stadium expenditures used $82.3 million of that $126.5. So, the 2013 changes added $44.2 million that remain in the reserve. Again, this came from tax increases that all parties (Governor, House, and Senate) proposed to use for other purposes. Thus, the 2013 changes set aside $44.2 million more of the general fund presumptively for the stadium. That money otherwise would have been available for any general fund purpose.

Fiscal year 2019 and following. The ongoing effect of the 2013 changes is the $44 million carryover. Future allocations are now driven by growth in lawful gambling taxes and the Minneapolis sales tax coming online in 2021.

To summarize some of the key ongoing effects:

• Amounts in the reserve will grow dramatically (to $360 billion by fiscal year 2025 in MMB’s latest projections).
• Under standard budgeting processes, the moneys are set aside for stadium uses. They do not show up in forecast surpluses or deficits for the general fund.
• The administration could spend them for stadium related purposes, subject to notice to and comment by a legislative commission. So far administrations have only used the moneys for amounts already appropriated by the 2012 stadium legislation. It is not legally limited to doing so.
• The money is not constitutionally dedicated; the legislature could use it for any purpose. But without legislative action it is trapped or sequestered in the reserve.
• The reserve increases the stadium’s cost to the general fund because it sets aside more lawful gambling tax revenues than needed to pay the bonds and stadium operating costs. That additional cost (set aside) grows significantly when the Minneapolis sales tax becomes available for stadium purposes. As noted in the previous bullet, the legislature could negate that at any time. But until it does, the reserve increases stadium costs by a fairly large amount. As an editorial aside, if its reality – that the reserve was not needed to finance the stadium and that it would increase the stadium’s public costs by 20% or more in later years –had been known to the 2012 legislature, it is unlikely the reserve would have been included in the bill.

The dog that didn’t bark

One of the functions of the reserve account calculations was to allow triggering of a suites rental tax if increased gaming revenues fell short of that needed to for state stadium costs. One could have objected to the 2013 changes because they made that more difficult by adding revenue to the calculations. The added revenue was not enough. The Dayton administration could have triggered the suites tax in 2015-16 but chose not to do so – it probably regarded the team contributions negotiated in 2012 as sufficient.

Note: the 2021 Senate omnibus tax bill repeals the authority to impose a suites tax. At this point, that authority is obsolete if it ever was more than a feel-good provision to appease those who thought the team and users should pay more. Its existence likely caused hand wringing by the Vikings after the stadium bonds were issued when MMB could have imposed it. It is all but meaningless now as far as I can tell.

Conclusion

The net result is that the state now has a goofy reserve account that captures more general fund revenues than needed to pay stadium bonds and for operations; current projections show those amounts will grow like Topsy potentially distorting state general fund budgeting. As will be discussed in the third post, I am hard pressed to come up with a reason for continuing the reserve or at least adding money to it.

This Federal Reserve Board discussion paper, Political Connections, Allocation of Stimulus Spending, and the Jobs Multiplier by Joonkyu Choi, Veronika Penciakova, Felipe Saffie, reaches some striking conclusions about the correlation between political contributions to state legislators, the likelihood of getting grants under the 2009 Obama stimulus legislation (ARRA), and their employment effects. Two authors are economists who work for the federal reserve system and the third for the University of Virginia.

Here’s the abstract which succinctly captures their research and findings:

Using American Recovery and Reinvestment Act (ARRA) data, we show that firms lever their political connections to win stimulus grants and public expenditure channeled through politically connected firms hinders job creation. We build a unique database that links campaign contributions and state legislative election outcomes to ARRA grant allocation. Using exogenous variation in political connections based on ex-post close elections held before ARRA, we causally show that politically connected firms are 64 percent more likely to secure a grant. Based on an instrumental variable approach, we also establish that state-level employment creation associated with grants channeled through politically connected firms is nil. Therefore, the impact of fiscal stimulus is not only determined by how much is spent, but also by how the expenditure is allocated across recipients.

Political Connections, Allocation of Stimulus Spending, and the Jobs Multiplier

The paper is timely with the enactment of the $1.9 trillion American Recovery Plan Act (ARPA), which is being partially sold as stimulus, and with Biden proposing spending another couple trillion on an “infrastructure” plan. If political connections (campaign contributions is their measure) yield materially more grants, that certainly raises ethical concerns. It’s even worse, if shoveling money to those firms yields suboptimal outcomes, which they define as stimulating employment (that is a secondary reason for infrastructure projects obviously – the main goal is to construct useful facilities).

So, how did they do empirical analysis to measure these effects and how much credibility should we put in the findings? I’m not competent to address the latter question, so I’ll punt on that beyond making a couple of observations. I am modestly skeptical but their findings are still concerning.

Structure of the research

There were two elements to their research:

1. Determining if “politically connected” firms got more grants because of their campaign contributions.
2. Measuring whether giving those firms more money affected the outcomes – i.e., the number of jobs created relative to the dollars granted to other firms.

What they found

Part 1: Do campaign contributions yield more grants?

The authors used data on grants to businesses from the 2009 Obama stimulus (ARRA) program to determine whether making more campaign contributions enabled firms to get more grants. Obviously, some sorts of sophisticated controls are necessary to avoid picking up random noise or spurious correlations, rather than likely causal relationships. They did that (or attempted to, anyway) using several techniques:

• They assembled a data base that combined ARRA’s data on the firms that received grants directly from states (“prime” vendors, not sub-vendors, or vendors getting contracts from prime vendors or grants from sub-grantees, such as cities) that they matched with campaign contributions made by the firms to legislators in the states making the grants. The campaign finance data was from 2006-08, the period immediately before ARRA was enacted, and was limited to contributions to legislators serving during 2009-10, when the grants were made. The campaign finance data was from National Institute of Money in Politics. They report that firms made 16% of campaign contributions (28% on a dollar-basis). That was a surprise to me, since I had thought businesses directly contributing was rare and more typically owners and employees contributed. But I have never been involved in campaign finance in any capacity and my knowledge is limited to Minnesota. They point out that campaign contributions are predominately made by small and medium sized businesses and those were also the firms that got most of the ARRA grants (see graph on page 10).
• To add a randomization factor – this is key to avoid measuring spurious correlation – they focused on a narrower category of campaign contributions, ones made to legislative candidates in races that turned out to be close (i.e., a 5% or smaller margin of victory) – about 10% of elections in the sample. Their assumption is that for such close elections, the result is effectively random or unpredictable and that will control for other factors (endogenous factors) that could skew the results.
• The dependent variable (what they are trying to measure the effects of) is the ratio of contributions made to winning candidates in close elections to total contributions made in close elections (pp. 12 -13). Firms whose contributions in close elections went 50% or more to winners are “treatment firms.” Thus, they’re measuring whether those firms were more likely to get ARRA grants than firms whose contribution recipients (again, only in close elections) did not win. To avoid distortions by comparing unlike firms, they created pairs of firms (i.e., a treatment and a nontreatment firm) based on the number of campaign contributions made in close races and by industry sector.
• They also added controls for the number of contributions made, firm size, and firm age.

Results. They found that the politically connected firms (i.e., those who contributed to more winning legislative candidates in close races) were 64% more likely to get an ARRA grant. Controlling for the size of the firm, whether it had its headquarters in the state, or how long it had been in existence did not change the regression results. To test for a placebo effect (essentially a correlation caused by something else), they checked and found the treatment firms did not get more ARRA grants in other states or as sub-vendors in the same state. They did several other statistical tests that I won’t summarize to rule out mismeasurement.

My take

I started reading this with a healthy dose of skepticism – it seemed to me that there is a lot of potential for unobserved factors, loosely correlated with making campaign contributions, driving the results. But re-reading and thinking about their research dispelled much of that skepticism. They were careful to develop techniques to screen out that risk. My one lingering concern is that elections in the 2006 to 2008 period (their sample) were Democratic wave elections. So, they may be picking up firms that tend to give more to Democratic candidates and that may not be as random as they think the results in close elections are. But that probably isn’t as much a factor/concern given the results in red states where the Democratic-friendly political environment just made Republican wins closer.

Part 2: Employment outcomes

Their next task was to determine whether skewing grants to firms that made fortuitous campaign contributions mattered – i.e., did it adversely affect the number of jobs created, the primary goal of the ARRA stimulus. To answer this question, the authors turned to an economics literature that I was ignorant of (even though I am familiar with some its authors).The literature analyzes the regional variation in the employment outcomes of the 2009 stimulus to see whether it was effective in stimulating job creation. Put another way, this literature tests whether states that got more money per capita created more jobs, when other relevant factors that differ by state are controlled for. The authors repeat this analysis but divide the grant allocations into those going to politically connected firms and to non-connected firms (again based on contributing to winning state legislative candidates in close elections).

To apply this approach, they use variations across states – essentially categorizing states by how much of their grants went to politically connected firms to test the results. There is quite a lot of variation in the state levels (no surprise) – see Figure 4, p. 21. That variation enables them to do their statistical analysis. The prior literature controlled for economic conditions (obviously, more ARRA grants were likely to go to states that were more severely affected by the Great Recession) and quality of state management (also may attract more grants by making better applications to the feds). That leaves them the task of randomly controlling for political factors; those too may be correlated with the likelihood of getting more grants (being endogenous) or other confounding factors.

They used two factors – whether a state allows corporations to make campaign contributions and the size of its legislature – to measure variation in states’ steering grants to political favorites. In their view, states with more legislators (per capita obviously) have more possible recipients to give to and (hope to) influence. Allowing corporate contributions evidences more opportunity for businesses (corporations anyway) to contribute.

Their analysis confirms the findings of the prior literature – i.e., that the stimulus grants increased employment (about 13 jobs per $1 million) – when the grants were made to firms without political connections. That was not true for states where more money went to politically connected firms. In their words:

In fact, according to our estimates, 21 percent more jobs were created or saved in a state like Pennsylvania, in which only 2.1 percent of ARRA spending went to politically connected firms, than in a state like Illinois, in which 22.7 percent of ARRA spending went to these firms.

Political Connections, Allocation of Stimulus Spending, and the Jobs Multiplier p. 27

This is consistent with other empirical literature that finds inefficiencies in government contracting with politically connected firms (another literature I was unaware of). They conduct several tests to validate that their findings are robust, controlling for various factors (industry composition, flexibility of labor, firm specific factors, etc.) that they think could skew the results.  None of those tests undercuts their confidence in the findings that:

[N]onpolitically connected firms create 13 jobs per million dollars, while the employment multiplier associated with grants to politically connected firms is not significantly different from zero.

Political Connections, Allocation of Stimulus Spending, and the Jobs Multiplier p. 3

My take

I have much more difficulty with the second part of their analysis, although my competence to judge it is thin. I have not read the background literature – either that assessing the effectiveness of the stimulus or on effect of political favoritism on the efficiency of government grants and contracting. But two factors give me pause:

• Their two variables strike me as perhaps tenuous as measures of the potential for campaign contributions to influence decisions. Yes, the law allowing corporate campaign contributions likely is somewhat reflective of a state’s political culture. But I assume most of those laws were enacted generations ago – probably during the Progressive Era at the beginning of the 20^th century. How reflective is that now of a state’s political culture? I get that they think the ability to contribute is the key thing – it’s all about opportunity. But my feeling is that a culture of accepting political steering of grants to political favorites is a big deal. But maybe stuff like that is persistent. Illinois has been corruption-friendly for years; nothing changes?  The size of the legislature variable also strikes me as problematic, since my instinct is that political power and influence is concentrated in a small number of legislators regardless of a body’s size. If that is so, it would not matter so much that contributors have more targets for their contributions and potential influence. Moreover, in Minnesota distributing federal grants is largely an executive branch function and my impression is that legislators have little influence over that, but maybe I am just oblivious or naïve. Practices and rules (regarding legisaltive involvement) obviously may differ. In any case, these are little more than instincts and suppositions on my part. My skepticism, perhaps, should be discounted by their finding of strong statistical correlations?
• I’m a bit more perplexed by their finding that grants given to politically connected firms had no employment effect. That does not seem to square with what I would expect based on the underlying economic theory of Keynesian multipliers. I had thought the whole theory was simply to pump money out into the economy via government spending that results in the hiring of workers who will, in turn, spend the wages and stimulate more employment. (Recall that Keynes famously suggested hiring workers to dig and fill holes would work.) Having absolutely no effect (“nil” is their term) implies the politically connected firms are somehow keeping most of the money from going into incremental spending (putting it in bank accounts, buying existing stocks and bonds, etc.). Okay, I get that is likely to happen at the margin – they do it with some of the grant money – but to have no effect at all? Yeah, yeah, I remember shoebox scandal in Illinois and yes, some of grants probably end up as kickbacks, bribes, etc. that go into shoeboxes or offshore bank accounts. But even so, their results are so surprisingly bad that they verge on the implausible to me. If they created one-third or one-quarter of the jobs, that would make more sense. Of course, measurement of this type is always imprecise.

In any case, the paper is interesting, and its findings are concerning. I’m not sure – assuming one is confident in their results – what policy solution could address the problem. It’s not practical for federal civil servants to hand out the grants and at the state level it is hard to imagine a way to devise a formulaic type mechanism for doing so (e.g., similar to using competitive bidding for procurement contracts). But grant making has never been something I have focused on, so there probably are innovative ways to help insulate the function from politics a bit.

I spent way too much time during my life as a legislative tax staffer trying to figure out what did and did not violate Grover Norquist’s “no-new-taxes” pledge. That pledge was signed onto by nearly all elected Republicans staring in the early 1990s. It hamstrung their ability to make sensible tax and budget policy decisions. Ultimately, some of them (e.g., Governor Pawlenty) got up the courage to look the other way on a few minor measures during very tight budgets (e.g., he agreed to partially close the FOC loophole in the corporate tax to raise a small amount of revenue).

In theory, the Norquist pledge allows revenue neutral tax changes. So, that should allow classic tax reforms that expand the base and make offsetting changes, such as cutting rates. Revenue neutral tax reforms require strong doses of political courage because their inherent dynamic creates concentrated groups of losers (from repealing deductions and credits) that benefit a diffuse group of winner (almost everybody benefits from rate cuts). Any politico knows that will typically not turn out well: an energized group of losers ends up hating you while a big group of beneficiaries doesn’t even know you helped them. To make matters worse, the political aura created by the pledge effectively demonized (in most elected Republican officials’ minds) tax changes that raise anybody’s taxes even if they didn’t raise net revenues. Of course, the most compelling reason to make tax changes is to provide revenues to pay for government services. After all, that is the function of taxes, which everyone understands. But that is precisely what Norquist’s pledge prohibits.

It is not just Republicans that make dumb campaign tax pledges. Joe Biden made a potentially even worse one during his campaign – that no one making less that $400,000 would have their taxes raised. Len Burman, a guy with Minnesota connections (he got his PhD from the U’s econ department), has a great TPC post that explains how Biden’s post – if he follows it – will thwart good tax policy and fulfilling some of his other promises.

On some levels, Biden’s is worse than the Norquist pledge because it blocks cleaning up senseless tax expenditures and prevents user-like taxes. For example, my view is that funding highway and road improvements should be done mainly with user-related taxes. The natural way to do that is with a hefty increase in the gas tax and some sort of alternative fee for EVs. That would have the environmental benefit of mildly discouraging fossil fuel use and linking beneficiaries of the spending with the payers. That is not permitted under the Biden pledge, as Burman’s post makes clear.

Instead Biden’s actual plan funds highways and road with corporate tax revenues and encourages EV use by subsidizing their purchases and charging stations. A gas tax increase would likely be more effective in encouraging EV use. I get that it is a hard political sell because people dislike obvious pain and opponents can easily demonize it. This TPC post points out: “The motor fuels tax has been frozen since 1993 and is the lowest in the OECD.  At 18.4 cents per gallon, it imposes a price of about $18 per ton of carbon dioxide emissions, far below the Administration’s own estimate that the social cost of carbon is around $51 per ton.” So, gas tax funding has the inherent advantage of user financing and is environmentally friendly (a classic Pigouvian tax). But the Biden tax pledge makes it a no-go.

[Note: After I wrote this but before posting it, Howard Gleckman makes a similar point that Biden’s pledge eliminates the best alternative for combatting climate change, a carbon tax. That’s a somewhat more complex issue. The general demonization of all taxes has effectively made an explicit carbon tax essentially a no-go, despite the preference of most economists – and me – for that approach.]

Len’s post is worth reading. I hope it is one campaign promise that Biden abandons. There are a host of changes that would be good to make in the federal tax structure on a revenue neutral basis or that may be the best way to raise revenues to pay for his initiatives, but neither can be done under pledge to immunize anyone making less than $400k/year.

As Len says, “Tax pledges are terrible policy and dubious politics. Politicians of both parties should pledge not to handcuff themselves by making stupid pledges.”

They, of course, stem from the age-old desire to convince people by offering them a free lunch. Human nature abhors a budget constraint; free stuff is compelling. Like all such ploys they are fundamentally based on misrepresentation or deception, which comes in two different flavors depending upon the party:

• Democratic flavor – we can tax somebody else (the “rich” or those making > $400k in this instance) to pay for all the government service we promise and that you want.
• Republican flavor – we can cut taxes without cutting government services that you want. This is slightly more complex but runs something like taxes are already high enough to finance all the government services you want (never mind that the federal government is running massive deficits and state taxes do not keep pace with population, inflation, and economic growth) and, in fact, are now paying for services you do not want. The latter is a deception because they are rarely forthcoming about what services they would cut and their failure to follow through implicitly says voters do want those services. Like the Democrats (with taxing the rich), Republicans in Minnesota try to justify their assertions about cutting spending by identifying programs that are perceived to not benefit their voters (e.g., LGA for center cities and similar).

Unlike Len, I am not convinced that either approach is bad (“dubious” in his terms) politics if the measure of good politics is electoral success. I find it difficult to argue with the political appeal of pitches the parties have now made for decades. Moreover, these “free lunch” pitches follow the approach many businesses successfully used to play on the foibles of human behavior (luring customers with “free” offers, highly misleading pricing, etc.) to sell their products and services. What I find incontestable is that they make for bad governance and they are sold based on some form of misrepresentation.

Of course, tax pledges are not binding contracts or laws. So, the real issue is how much they actually change politicians’ behavior. How willing are they to just look the other way or to say circumstances have changed since the campaign? We know that most Republicans take their pledges very seriously – probably thanks to what Pat Buchanan did to Bush 41 after he abandoned his “read my lips” pledge. The fealty of Dems to their “we’ll only tax high income folks” is somewhat less clear, although we have evidence many of them pretty much honor it (ignoring the economic incidence of corporate taxes partially falling on consumers and recasting sin taxes as something else). Here’s hoping they’re willing to move on or forget their pledges, but I am not optimistic.