Last week, in a comment on this blog, the highwayman stated, “road users only cover about 20% their costs directly and the rest of the funding comes mostly from income & property taxes.” I don’t know where he gets this information, but it is not credible.
The National Association of Railway Passengers (NARP) claims that “41% of the $133 billion spent on highways came from payments other than the gas tax, tolls, and vehicle taxes and fees.” In particular, NARP counts the proceeds of bond sales and interest on savings as money that comes from sources other than user fees. But how do the bonds get paid back? Mostly out of user fees. Where does the interest come from? Mostly from savings of unspent user fees.
What is the truth about highway subsidies? And what ought to be done about them? The Antiplanner has addressed this topic before, but (judging from the comments) it may have escaped the attention of some readers and I confess that I may not have rigorously covered this subject.
The best source of information on this issue is the finance section of the U.S. Department of Transportation’s annual report on highway statistics. Table HF10 summarizes where the money comes from and where it goes.
Line 15 shows that federal, state, and local governments collected $116.4 billion from highway users — gas taxes, tolls, vehicle registration fees, and similar fees collected for the specific purposes of providing money for highways. General sales taxes on cars and other materials are not included. Of this amount, $3.2 billion is spent on collection expenses (line 18), leaving $113.2 billion for roads.
Except it doesn’t all go for roads, as $10.5 billion was diverted to mass transit and $8.8 billion was diverted to other purposes. In addition, for several years in the 1990s, 4.3 cents of the federal gas tax were diverted to “deficit reduction,” and they did not show up on table HF10 at all. I count these all as highway user fees because people paid them in the course of using the highways, not as a general tax (like a sales tax on all goods).
In any case, we have $113.2 billion in highway user fees. Line 32 shows that another $44.5 billion in property taxes and general funds also went to highways. These are the subsidies. My simple method of calculating subsidies is to subtract the $19.3 billion in diversions to transit and non-transportation purposes from this $44.5 billion to get the net subsidies — in this case, $25.2 billion in 2006.
Since the entire amount spent on roads was $161.1 billion (line 66), subsidies were about 16 percent of the total. This is more than usual — historically, the average has been about 12 percent, but it has gone up in the past couple of years. By comparison, transit gets about 73 percent of its funds from tax dollars and only 27 percent from user fares.
Another way of looking at it is to divide the $25.2 billion subsidy by the 4.9 trillion passenger miles of travel shown in table VM1. The result is about a half penny a passenger mile. This is far smaller than the subsidies to transit, which averaged about 61 cents per passenger mile in 2006.
This is the way I have previously calculated highway subsidies, but I confess that there is a slight flaw in this analysis. It presumes that all of the bonds (shown in line 35 of table HM10) will be repaid out of user fees, not taxes, and all of the interest (shown in line 33) comes from savings of user fees, not taxes. This is mostly, but not entirely, true.
Most of the bonds and a lot of the investment income is at the state level, and the states get 88 percent of their income from user fees (even more if you count federal distributions to the states). Table SF3, for example, shows that the states with the biggest bond sales — New Jersey and Texas — got none of their highway money from general funds.
But let’s ignore all that and just look at user fees and taxes. If users paid $113.2 billion and $44.5 billion came from taxes, then 28 percent of highway costs came from taxes and only 72 percent from user fees. That is higher than I’ve stated in the past. But it still leads to the conclusion that highway subsidies are less than a penny per passenger mile. I think a close scrutiny of the bonds and interest would show that the actual user fee contribution is higher than 72 percent, but no higher than 84 percent.
Not only are taxes used to subsidize 73 percent of transit but no more than 28 percent of highways, this has been true since at least 1970, the earliest year for which we have much transit data available. In most years since then, the total number of dollars spent subsidizing transit actually exceeded the dollars subsidizing highways — even though roads carry 100 times as many people and far more freight than transit. So any idea that there is some “backlog” of transit work needed to “catch up” with historic highway subsidies should be forgotten. If after nearly 40 years of huge subsidies transit’s market share of urban travel is still only about 1.5 percent, more subsidies are not the answer.
I go through this only to dispel any residual claims that road subsidies amount to 41 percent or 80 percent or whatever higher percentage of the cost you might hear. This doesn’t mean the Antiplanner approves of these subsidies. My preference is to turn all roads, or as many as is feasible, into toll roads. Charge market rates for the tolls. Build new roads only where tolls justify their construction. Earn profits on the roads that you can. Reduce maintenance or other costs on roads that lose money; close them if that is the most expedient thing to do.
That is not the argument we hear coming from planners and planning advocates. Instead, they say things like, “all transportation is subsidized, so ignore the subsidies that are going into rail transit and Amtrak.” Or, “road subsidies tilted the balance against transit and forced Americans to drive when they might have chosen other options.” In fact, they simply want more subsidies for their favored mode of travel.
Let’s just end all the subsidies and let the transportation chips fall where they may. Or, if you think some people need subsidies because they are poor, disabled, or otherwise disadvantaged, give them the subsidies and let them decide how to use them rather than subsidizing some transit bureaucracy or construction industry. If ending the subsidies means more rail transit, great! I love trains. If, as I suspect, it means less, we’ll still have plenty of privately funded historic rail and trolley museums and tourist lines.
Speaking of which, the Midcontinent Railway Museum, the nation’s premiere museum specializing in operating rail equipment of the 1880-1916 era, was hit hard by the Midwest floods and needs the support of those who appreciate rail history. The Antiplanner sent in a donation and encourages other train lovers to do the same.