Eno Foundation's TRANPORTATION QUARTERLY (Summer 2003) carries an exceptionally silly piece called "The inefficiency of toll collection as a means of taxation: evidence from the Garden State Parkway" by Jonathan R Peters and Jonathan K Kramer, a couple of junior northeast academics. They start by alleging...MORE
Eno Foundation's TRANPORTATION QUARTERLY (Summer 2003) carries an exceptionally silly piece called "The inefficiency of toll collection as a means of taxation: evidence from the Garden State Parkway" by Jonathan R Peters and Jonathan K Kramer, a couple of junior northeast academics. They start by alleging that most tollroads are monopolies, suggesting they exploit market power to extract excessive profits from motorists. They cite as examples the Pennsylvania Tpk and Garden State Pwy (GSP) which they claim occupy "monopoly controlled corridors." Can't these guys read a map? Have they driven the corridors?
In fact the PA Tpk is paralleled very closely by US-30 most of its length and has interstates I-68 and I-80 on either side of it that compete for longdistance traffic. Similarly the GSP is followed for much of its length by US-9.
Now of course the tollroads provide a faster less interrupted trip, so there isn't perfect competition. If the competition were a free road providing an identical level of service – a freeway right alongside – no one would use the tollroad.
The junior profs next say "if the goal of instituting a toll is to reduce pollution, is a toll more effective than a gasoline tax..." Who ever suggested pollution reduction was an objective of tolling? In eight years of following the industry I've never heard anyone say that. Talk about a strawman.
They next cite a study about the impact of tolls being "regressive." But relative to what? And so what? If every service was funded by "progressive" taxes the nation's entrepreneurs would all go away, the income tax rates would be so high. They'd go offshore to unprogressively taxed places, or, onshore they'd devote their talents to tax dodges rather than productivity. So some stuff has to be funded by "regressive" means. Why not roads?
Little work, they complain, has been done "to quantify the impact of toll collection on society as a whole" and these intrepid researchers set out to remedy that deficiency with one paper generalizing from some numbers conjured up from one tollroad during one small portion of its existence. This is a bit like saying we don't know the full impact of lung cancer on society, finding one man with lung cancer, getting some data about him, for say, a month of his disease, and saying that writ large that is the impact of lung cancer on society.
Their results, they say "show that toll collection is a very inefficient means of taxation." Sure, but tolls aren't taxation. They are a price, a charge for use of a facility which was financed in the capital markets because gas tax money wasn't available.
There is no discussion of whether the GSP is a typical tollroad, what variability there is in the costs and revenues of different pikes, whether the years from which they extract their data are average years.
They just assume you can get one year's data from one tollroad and generalize to the industry.
You can't. And, sure enough, the year they chose for their GSP data, 2000, was a very unusual year. Toll collection costs were very high because E-ZPass was just starting to be implemented and they were having trouble implementing it. And of course no economies were yet being reaped in manual collection and coin machine costs because E-ZPass was only just getting operational. Also electronic tolling was atypically difficult to implement on the GSP. Several of its major mainline plazas have dreadful curving belly-outs that make it awkward to do good advance signing to get drivers in lane, because it is tough designing a sign gantry that spans 30 toll lanes. That's unusual, and it made for unusual difficulty, delay and cost at the GSP. ET hasn't caught on as fast on the GSP as elsewhere either. As a result, its benefits have not been as great as most other places. But they are working on it, and will get it right eventually. And costs will come down.
But no doubt the GSP's costs are high. They are a union shop. And they have barrier plazas with serious queueing. And their toll rates are low, so toll collection cost as a proportion of revenue is unusually high at the GSP. But all that says zilch, zip and zero about the costs of toll collection in general around America or at other times than during a difficult implementation of electronic tolling on one pike.
The pair conjure up a whole lot of numbers on time savings and pollution reductions which they imagine might be attained by ending toll collection. It is all nonsense. Here it is in black and white: "The estimates are based on the 11 major toll barriers, because if they were not present, traffic would flow freely." (p28)
Anyone who bothers to read a little New Jersey history knows that if the toll barriers were not there on the GSP, the GSP would not be there. The traffic would be traveling stop-and-go on US-9 and other slow signalized arterials where there are interruptions to traffic flow far more numerous than toll plazas.
If now in 2003 our pair of profs managed to use their persuasive powers on the state legislature to pony up a whole big stream of gas tax money against which the tolls could be ended, then the plazas could be taken down. Good luck. But they are wrong again in assuming there would then be free flow.
The GSP already, with tolls and the metering effect of the plazas, has separate beyond-the-plaza congestion points, especially just north of the Raritan river, but in several other places as well. But the elimination of tolls would first eliminate the plaza metering effect which protects many downstream points from congestion. And second as a free road it would attract a good bit extra traffic. These two changes - no metering and the lower price - would produce many new congestion clots on the GSP. It would NOT bring free flow.
Guys, if eliminating tolls guaranteed free flow the toll-free urban freeways of this country wouldn't be suffering congestion. We wouldn't have all those multicolored realtime website maps of different metro area speeds, radio reporters patrolling the freeways by helicopter, the Schrank-Lomax annual congestion books, or Tony Downs advising the motorists of America to get used to stop-and-go, relax, and get a better car stereo.
Finally, of course, free flow CAN be achieved at the mainline barrier plazas by rebuilding the central sections of the plazas specifically for highway speed operation with transponders - as done on several dozen tollroads that have escaped these guys attention. And indeed: as the Parkway itself is planning in its retrofit of open road tolling. Our prof pair apparently have never heard of the million motorists a day plus now paying tolls on roads where there is simply no cash collection at all - 407-ETR, CityLink, 91X, CLCC, TIH etc - where no one at all stops to pay.
This point has not penetrated these thick heads of academia: freeflow at the toll collection points, indeed the complete banishment of all cash collection at the road, is now perfectly compatible with tolling. The whole assumption underpinning their TQ paper methodology - that toll collection requires delaying traffic - just ain't so anymore. It's all been swept away by transponders and cameras that function at 100mph.
Unfortunately the unmitigated nonsense and uninformed silliness of this article will be TQ's last word on tolling. We offered a corrective commentary, but heard Eno have already gone into production of their final edition.
PS Searching desperately for something positive to say about this article, I came up with this. It is clearly written. Unlike many of their ilk, these two don't bury their work in obfuscations, jargon, or rambling circumnavigations around the periphery of the subject. They focus and go straight in. They write well. Unfortunately their very clarity makes crystal clear their poor methodology and profound ignorance. TRnews 2003-07-19
