This is another installment in my effort to think through the debate over highway funding. WARNING: THIS IS LONG AND BORING; IF YOU DECIDE TO PROCEED, YOU MAY WISH TO HAVE A MUG OF ESPRESSO HANDY.
There appears to be a bipartisan consensus that the state needs more money for highways (and for transit, although Republicans see little need for that, I’d guess), raising an obvious question:
Should the gas tax or tab fees (registration tax) be increased or should revenues from general fund taxes be reallocated to highways? The state has been consistently running general fund surpluses; why not use some of that money?
This question is complicated by Minnesota’s long tradition of using dedicated funding for highways and to a lesser extent and for a shorter period transit. It has done that mainly through constitutional dedications for highways and roads, starting in the 1920s and modified by several constitutional amendments over the years. Constitutional dedications require the legislature to use gas tax and the motor vehicle registration tax (tab fees) revenues only for state highways and for aids to counties and cities for local highways, streets, and roads. Since 2008, the sales tax on motor vehicle purchases has been dedicated to transportation (not just highways, at least 40% must go for transit). The legislature has occasionally supplemented the constitutional dedication with statutory dedications; currently some of the sales tax on auto parts is dedicated by statute to highways and roads. Since it is a statutory dedication that the legislature could always change; the Republicans, as a result, want to write it into the constitution making sure a future legislature can’t.
This practice of using dedicated funds for transportation channels much of the debate into questions of whether the dedicated taxes should be raised or if general fund taxes/money should be redirected to transportation. Obviously, Republicans with their strong tax aversion routinely advocate for redirecting existing taxes. Democrats, by contrast, typically advocate raising the already dedicated taxes – e.g., Governor Walz and House Democrats both supported large gas tax increases in the 2019 legislative session – although they are conflicted by the fact that the dedicated taxes are regressive, violating their norm of favoring progressive taxation.
Given Minnesota’s long tradition of funding transportation heavily through dedicated funds, I see little value in spending time on the merits of that. As an aside, in the lead up to consideration of the Legacy funding constitutional amendment in 2008, I wrote a publication on earmarking taxes, which I subsequently updated in 2015. It discusses these issues in some detail, so I would also be needlessly re-ploughing old ground. In any case, it is a given that almost all of state highway funding in Minnesota will be done through dedicated funds.1
The more important question is whether any increase in funding should come from user fees and benefit taxes or from general tax revenues (e.g., by earmarking existing taxes for highways). I think a strong case can made for providing any increased state funding for highways and roads through user charges or benefit taxes, rather than taking more general revenue taxes and using them for highways. This post explains my thinking.2
As an initial matter, it is useful to state how I think about the difference between a user charge or benefit tax and general tax revenues. Here, I’m talking about functional, economic, or practical distinctions, not legal distinctions. Any tax can legally be earmarked or dedicated and, then, it becomes part of the dedicated fund. But that legal status does not make it either a user charge or benefit tax, even if the tax is somehow closely or loosely related to transportation or whatever function the dedication relates to.
A user charge or a benefit tax should satisfy two criteria:
- It needs to be over and above the taxes that generally apply to most similar activities for general government funding and
- The amount should be closely correlated to consumption of or benefit from the funded government service.
A true user charge would vary by the amount of the service consumed or used; if you use more, you pay more and ideally would be voluntary. By contrast, a benefit tax just needs to be collected from the group of individuals or entities that particularly benefit from the service.
Note the effects of this definition: Dedicating part of the proceeds of a general revenue tax does not qualify. For example, taking the sales taxes paid by trucking companies and dedicating it to highways and roads is not a benefit tax or user charge. Those taxes are levied on all businesses and individuals to pay for the general cost of government. It flunks criteria #1 and is not charging users or beneficiaries of a government service additional amounts because of that status. This distinction explains why the sales tax on purchases of vehicles or auto part are neither user charges nor benefit taxes. They are just taking or appropriating general tax revenues and legally earmarking them for transportation.
In a separate post, I go through Minnesota’s funding of highways and roads and attempt very roughly to identify how much comes from user charges and benefit taxes versus general government revenues. In broad strokes, the motor vehicle registration tax is a benefit tax and a portion of the gas tax is a user charge. The rest of highway and road funding comes from state or local general revenues – property taxes or state aid that locals opt to use for roads or general sales taxes that have been legally earmarked for highways. Based on my calculations, about 59% of highway and road costs are paid with general revenues. Somebody with better analytical skills, access to and understanding of the data, and more time could estimate a more accurate number; mine is just a back-of-the-envelope effort. I do think it is a conservative estimate of the general revenue share; the actual share likely is higher.
So, why do I think increasing user charges (lets be transparent, the gas tax) is the best way to provide more state funding for highways and roads? I have two prime rationales which are related to each other, in addition to a third, more minor but not inconsequential, reason for liking a gas tax increase. After I describe why I favor a gas tax, I’ll explain why I think the common objections to or concerns about the gas tax should not disqualify it as the principal funding source for additional highway funding:
First, basic economics – it is consistent with fundamental tax principles and helps a market economy function better. The common metrics public finance economists use to evaluate a tax are equity, efficiency, and simplicity. (Unlike the public and most elected officials, those are also the criteria I use in judging whether a tax proposal is sensible. I’ll get to political feasibility issues a little bit later.) My premise in applying these principles is that the gas tax is a rough user charge, a characteristic widely recognized by tax policy experts. It is an effective user charge because the amount of tax paid is strongly correlated with how much one uses highways and roads – the more gas purchased for highway vehicles, the more miles that are driven. (It’s important to also note that the tax is also borne by those who do not own vehicles or purchase gas, because its burden is passed along in the transportation costs embedded in the goods and services that they do purchase. This is actually a large portion of the burden of the tax – about one-third according to the Minnesota Tax Incidence Study.) It’s true that that correlation is not perfect, because of variation in fuel efficiency of vehicles, more all electric vehicles, and the road damage done by heavy vehicles (the latter two concerns I discuss below). But fundamentally it is a workable way to charge for relative road use by most. Moreover, there is a sort of default rule that when user charges are practical and easy to impose, that is the preferred way to finance a government service, unless doing so would undercut the fundamental purpose of the government program. In any case, this is how I apply those three basic tax policy principles to the gas tax:
- It’s obvious the gas tax scores well under the simplicity principle: the tax has been around forever, is easily understood (cents/gallon used for roads – what could be clearer?), is easy to collect from a relative handful of taxpayers (petroleum distributors), and one main compliance issue (the exemption of diesel fuel used off road) was fixed 30+ years ago with the innovation of dyed fuel.
- The tax increases efficiency because it encourages people to make economic decisions (where to live, what car to buy, etc.) that reflect the costs of constructing and maintaining public roads. The tax makes the market work better – the price of gas (and transportation costs more generally) will reflect more closely the public’s costs to build and maintain roads. This will flow through to decisions about where to live (how long/expensive a commute will I have?), what vehicle to buy, and so forth. If you, as I do, think the market is the best way for a society to decide how to allocate goods and services, a user charge for public highways is the way to go. And the gas tax is now the easiest way to do that.
- On the surface, it might appear that the equity principle is a harder nut to crack because the tax is regressive, the metric most often used in evaluating tax fairness. I have two basic responses to that concern. First, user charges, like the purchase of private goods and services, are fair because you’re paying for what you use. Here, it makes sense for the government to be the service provider because we want to provide open and easy access to the road network and private roads are not practicable. Putting aside that the government is delivering the service, charging for highways isn’t really all that different than individuals buying food, housing or other essential items in the private market, all of which have a regressive distribution (lower income folks spend higher percentages of their incomes on consumption). Second, the distributional effects are not that different than a general consumption tax, like the sales tax which is typically presented as the prime alternative funding mechanism. (The Minnesota Tax Incidence Study reports a Suits Index for the gas tax that is about one-third lower or more regressive than for the general sales tax.) And if regressivity is a killing concern, it can be addressed through supplemental mechanisms (see below), rather than forgoing a good funding source altogether.
Second, political economy considerations – relying on user charges could help make decisions that get beyond the pro- and anti-government budget debate we now seem to be mired in. Much of the discussion under the previous bullet represents conventional wisdom in the policy community. The views I’m expressing here are not but reflect my experience working in the legislature over the last 15+ years and how I think the budget and tax policy debate in the Minnesota legislature has been hijacked by the polarization and rote repetition of little more than partisan talking points. During that period, I found that the budget debate has devolved into largely meaningless sloganeering on the size of government and whether (GOP mantra) taxes are driving businesses, investment, and people out of Minnesota or whether (DFL mantra) the tax structure is allowing the affluent to avoid paying their fair share.
The Republicans have largely become a party whose primary (almost exclusive) budget policy is to propose tax cuts. There are a few areas in which they want to spend more money (i.e., beyond current services levels), but they can always propose reallocating money away from other programs with heavy DFL or urban constituencies to do so. That’s fine, but it also creates a “free lunch” element to their arguments: they’re for more limited government and tax cuts, except when the spending is popular with their constituencies, in which case they say the spending increases can easily be accommodated by reducing spending on DFL/urban type programs (center city LGA being the obvious target). What’s been unclear to me, because the GOP has never controlled the governorship, House, and Senate when I was around, is whether that is just a negotiating ploy or whether they really would do it.
In any case, transportation is one of the areas – I assume because it is a core government function and because it has a heavy rural focus (the GOP is fundamentally a rural and exurban party in Minnesota) – in which the Republicans do advocate for spending more (at least relative to what the inelastic dedicated funding sources produce). Their natural response is to make a sort of free lunch argument – sure the dedicated funding sources are not generating enough, but we can easily reallocate general fund sources to make up the difference. And to be consistent with Minnesota’s earmarking practices, we’ll do that permanently and submit it to the voters to enshrine in the constitution.3
To me these arguments undercut the debate over how much more to spend on highways, given our history of using dedicated funding, a substantial portion of which comes from de facto user charges and benefit taxes. A primary benefit of relying on user charges and benefit taxes is that increased spending comes at the price of paying more in user charges and benefit taxes (i.e., increasing the gas tax or tab fees). That is a more honest frame of reference for the public to consider how much they want more spending on highways and roads. Permanently reallocating general fund revenues (especially by constitutional amendment) reflects a judgment that the general fund does not need all its current revenues over the long run (not just when we have a surplus, like we do now). I don’t think there is a good case to be made for that – the median Minnesota voter has too much of a taste for government for that to be true.
My bottom line: Linking spending and taxing directly, as Minnesota traditionally has done for highway spending, is good political economy. I would limit this practice to areas where the linkage is relatively easy to implement and clear, such as highways. (I’m hard put to think of many other big areas.) The long history of that linkage working (until the 1990s) is a compelling reason for sticking with it. Maintaining the linkage tells voters that if they advocate for or expect more highway spending, the consequences will be paying more gas tax or higher tab fees. That is likely to lead, in my view, to better decisions about how much to spend.
Third, climate change considerations support increasing the gas tax; it is a mini-carbon tax or, at least, does not encourage more carbon emissions, as financing more highway expenditures with general revenue sources would. The gas tax is a carbon tax, since it is imposed on the use of fossil fuels. Of course, its low level and application only to highway fuels makes it a relatively minor tax. The fact that its revenues are used for highway construction and maintenance (spending which stimulates carbon emissions by encouraging more road use) makes its effects ambiguous. But compared to using general revenue taxes for roads, such as dedicating more general sales taxes, a gas tax is preferable if a secondary policy goal is to discourage carbon emissions.4
My responses to the common objections that are made to gas tax increases. Many arguments against increasing the gas tax are typically made during legislative debates. In my judgment, none of them, singly or together, are sufficient to overcome the case for the gas tax. The main arguments that I have heard and how I would respond are:
It’s regressive, unfair to the poor (the prime Democratic objection). The gas tax is more regressive than the sales tax. But contrary to the apparent DFL mode for evaluating taxes, that should not be the only criteria to use in deciding among taxes. Its other advantages, in my opinion, outweigh that disadvantage. User charges meet a very basic element of fairness – expecting people to pay for costs they impose or what they use. Moreover, other more progressive taxes in the general fund portfolio (i.e., the individual income tax) help to offset that disadvantage. Not every tax needs to be progressive if it has other attractive attributes.
A low-income credit, either tied specifically to the gas tax or as part of a more comprehensive credit to offset other regressive taxes, could be adopted to address this concern. The 2008 increase in the tax bill did precisely this to gain a handful of votes (primarily Tom Rukavina’s but others as well) necessary for the required supermajority to override the governor’s veto. The provision lasted only one tax year, however, as Governor Pawlenty and some DFL legislators united to repeal it and use the revenues for other purposes.5 The 2008 credit was poorly designed, and its effects were very modest at best. A better approach probably would be to convert the property tax refund into a comprehensive low-income tax credit/refund to offset a portion of the average property, sales, and excise taxes borne by low-income households. That has been discussed with some regularity and elected officials have shown little interest.
We have plenty of general fund revenues, so we can simply use more of them for a core government function like highways and roads (the prime Republican objection). I outlined my main reasons for rejecting this above. I would add that given the GOP’s persistent advocacy for tax cuts in all circumstances (even in 2011 with a $6 billion difference between projected revenues and expenditures!), they should recognize that reallocating general fund money to highway really reduces their ability to cut general fund taxes. At least that would seem to be the logic to me, and it undercuts their implicit “it won’t cost you anything” argument for using general fund revenues.
Trucks damage the roads more, but don’t use enough more fuel to offset that damage. This statement is (I assume) true and is a mild objection to the present excise tax; a more accurate user charge on heavy trucks and similar vehicles would be scaled to recover more fully the costs they impose on the road system. Because heavy vehicles are not very fuel efficient, they do pay more per mile in gas tax, but likely not enough to pay for the damage. (I have not attempted to research the magnitude of the difference, and I know a large share of the cost of Minnesota road maintenance is driven by our harsh climate, which subjects concrete and asphalt to breakup from frequent freezing and thawing.)
A simple fix would be to impose a higher rate of tax on diesel fuel, since almost all heavy vehicles run on diesel and very few cars do.6 Some states already do that. I’m not sure why Minnesota does not or if it has ever considered it. I’m sure the trucking and business lobbies would oppose it and that alone may make it politically infeasible. In any case, the objection of too favorable treatment of heavy vehicles is not a good argument for not continuing to use the gas tax as a cornerstone of our highway finance system and to raise any needed new revenues from it, rather than using general revenues. It’s a minor objection that could be easily fixed, if policy makers think it is a serious flaw and have the will to do so.
Electric vehicles (EVs) don’t pay tax and they comprise a growing share of the fleet of highway users. This statement is also true, but I don’t think it is a good reason for not using the gas tax as the prime source of added highway revenue. First, EVs comprise a trivial portion of the fleet using Minnesota highways and roads. I do not know what percentage EVs are of the fleet, but they are less than 1% of new vehicle purchases in Minnesota. That a very small share of the fleet uses electricity as power does not undercut that the gas tax still functions very well as an easy-to-administer, user charge on 99%+ of Minnesota vehicles.
Second, the government is trying to encourage purchase of EVs for environmental reasons by offering tax credits against their purchase price and by imposing fleet mileage requirements on automakers. A better way to encourage EV adoption and use is simply to push the price of petroleum-based fuels higher by raising the gas tax, rather than to push down the price of EVs with tax credits or by regulating what automakers can seller. Doing that targets the incentive to those we would most want to incent, i.e., heavy users of petroleum fuels, not those who have the income/wealth and the predilection to buy a Tesla (for status, feel-good green vibes, or whatever reason).
Ultimately, a system for imposing a user charge on EVs will need to be designed and implemented, because they will become a large part of the fleet. But that likely will take some time to occur and the surcharge on the tab fee that applies now helps somewhat make EVs pay their fair share. When the EV fleet becomes large enough, I assume it will be feasible to impose some sort of mileage charge on EVs, administered using a technological solution (GPS transponder attached to every EV?). Because EVs are all relatively new and expensive vehicles, it seems easier to implement a mileage charge for them, as compared with implementing such a user charge for all vehicles. The latter (ditching the gas tax in favor of a mileage fee for all vehicles) has always struck me as pie-in-sky and/or a clever way to thwart the sensible policy of just raising the gas tax.
The gas tax is unfair to owners of gas guzzlers, particularly those who are poor and/or don’t drive much (e.g., compared with high income owners of hybrids who drive a lot). This is essentially an argument that the gas tax is a flawed user charge or another version of the regressivity argument. My general response is that the tax is a pretty good user charge and falling for arguments like this allows the perfect to be the enemy of the good.
The gas tax is unfair to Greater Minnesota. This objection is based on the reality that a resident outside of the 7-county metro area typically drives more miles than a metro area resident – both because of geography and because the lack of density limits transit options outside the Twin Cities. I have two responses to this – one is politically unpalatable but reflects reality. Geography makes it more expensive to construct and maintain a road system in Greater Minnesota, not just for residents to drive and pay for gas. That reality must be reflected somewhat in how much residents of the region pay in tax; they can’t expect their metro aunts and uncles to neutralize this reality can they?
The second response is that a look at the numbers reveals that the situation is not that bleak. Yes, more gas tax is collected in the 80 non-metro counties (about 52.5%) than in the 7 metro counties. (For context, the 7 metro counties pay about 54% of the other state sources of highway finance revenues – tab fees and motor vehicle sales tax – counterbalancing that somewhat.) That is a relatively small difference. But the more important fact is that even more of the spending goes to the nonmetro area. About 40% of the state highway revenues are paid as state aid to local government. This aid goes heavily (about 65%) to the nonmetro counties (see the graph below). These numbers are from the House Research publication, Major State Aids and Taxes and are for 2015.
I could not find easy access to data on the geographic distribution of direct state highway spending; I’m sure that the metro/nonmetro breakdown varies from year to year. There are many more miles of state highway in the nonmetro counties, but maintenance and construction cost/mile is undoubtedly higher in the metro. Over the long run, I’d guess that number of miles outweighs the higher metro costs. (Maybe MNDOT puts out this type of breakdown, but I couldn’t quickly find it.)
The bottom line is that residents of Greater Minnesota enjoy a balance of payments that provides them comfortably more in spending than they pay in gas taxes. The balance payments is even more favorable when tab fees and motor vehicle sales taxes are taken into account. It’s unclear to me how this is an unfair situation. Of course, politicians representing rural areas can easily conjure up arguments with superficial appeal – typically focusing on what a rural voter would consider to be frivolous or wasteful general fund spending that occurs in the metro area. This “what-aboutism” muddles the issue in my view and is a prime reason why I favor a user charge that as much as possible links beneficiaries and payers, as I suggest above.
A per mile user fee (administered with transponders on every vehicle) would allow imposing a better more accurate user charge to replace or supplement the gas tax. I discussed this above under EVs. It just doesn’t seem very practical at this point to use as a replacement for the gas tax; I could see it as a supplement for EVs and hybrids along with a gas tax increase. It is just the latest shiny new object distracting from tried-and-true financing policies. In the long run (10 to 20 years), some sort of solution like this will need to be implemented and I think it will be driven by how fast we convert to EVs and driverless vehicles. Of course, there is a very good chance I’m wrong about all of this. A 2016 article, Jerome Dumortier, Fengxiu Zhang, and John Marron, State and federal fuel taxes: The road ahead for U.S. infrastructure funding, makes the case that the need for some sort of funding source beside the gas tax is coming much sooner than I think. It also provides evidence as to the effects of the lack of indexing of the gas tax that I like harp on.
I’m skeptical of how practical (including politically) a mileage charge is, based on the little I know about it. Minnesota would be the first state to adopt one (other than as a pilot project), which says a lot. This piece (Information Technology & Innovation Foundation, A Policymaker’s Guide to Road User Charges), makes the case for Congress enacting a national user charge now. Besides my skepticism about the practicalities and politics, I strongly prefer the gas tax because of my perception that it will better help address climate change. Of course, a mileage charge system would be better than dedicating part of the general sales tax on that score.
Notes
- I had thought no one read my earmarking publication, based on the very few hits it got on the House Research website, so I was surprised in 2018 when the North Carolina governor’s office scheduled a conference call with me to discuss earmarking. Apparently NASBO referred them to me based on the publication, suggesting someone must have read it. ↩︎
- I think it is fine for locals to use and increase property taxes for streets and roads, since there is clear benefit linkage between property ownership and construction and maintenance of roads. That seems better than the apparent growing urge, led by Duluth, to use local sales taxes. St Paul apparently wants to follow Duluth’s lead. Here, I’m confining my discussion to state funding issues, although increasing state funding through the highway user trust fund materially would reduce some of the pressures on local financing because it automatically yields more city and county aid. ↩︎
- I’m sure the voters would approve, just as they did in 2006 with the motor vehicle sales tax, because there is a strong political constituency for highway spending. They would probably approve a tax increase if submitted – given how voters responded in 2008 to the Legacy Amendment, easily approving a tax increase for what I perceive to be less popular spending. ↩︎
- I don’t want to underestimate the effects of maintaining a robust gas tax on carbon emissions. Based on my calculations if the gas tax had been indexed for inflation when it was increased in 1988 to 20 cents/gallon, the state would have collected about $6.5 billion more revenue (assuming little or no behavioral response to the higher tax rates). Had it done so, it is reasonable to conclude that it would have caused some behavioral response (less driving or a more efficient fleet) that would have lowered carbon emissions somewhat. ↩︎
- To me that suggests that the concerns about its regressivity are not really strongly held. ↩︎
- Diesel vehicles typically generate more pollution, so a differential rate might be justified on that basis as well. ↩︎
SALT Speak - Joel Michael's Blog 