This toll should really take a toll.
Most discussion of the current congestion pricing program in New York assumes that automatic tolls will be between $13 and $15 (round trip), with a lesser toll perhaps as low as $4 or $5, for off-peak rides. That’s according to Streetsblog contributor Charles Komanoff, who has been studying and modeling congestion tolls since the turn of this century, and believes that a $13 toll would be the bare minimum for the state to meet its annual revenue requirement. of $1 billion prescribed by law.
That $13 coincides with the tolls currently charged by the Metropolitan Transportation Authority on its six major bridges and tunnels. And, because the $13-$15 range has dominated congestion pricing discourse for years, it has created its own sense of political inevitability. Significantly higher car tolls are easily dismissed as politically unfeasible.
A congestion charge of $13 or $15 might be all politicians think they can get, but such a low charge will not maximize the net societal benefit conferred by congestion pricing. In fact, the toll must be much, much higher.
I determined this using Komanoff’s Balanced Transportation Analyzer modeling tool.
As Streetsblog readers know, the BTA is an open and freely available model for analyzing how different tolls affect key outcomes of congestion pricing in New York – time savings for drivers, revenue that can improve service transit (and therefore create time savings for straps), environmental benefits, etc.
When these advantages are offset by the disadvantages of tolling – administrative costs, lost amenities for drivers and the costs of tolling itself – the result is the total net societal benefit.
Komanoff said he never ran the BTA model for a series of possible toll levels to track the rise and fall of those outflows. So, using the BTA, I ran the numbers with a congestion charge of $0 to $100 in $5 increments. Here is what I found:
The first thing to notice is that as the toll increases, the net benefit to society increases, plateaus, and then begins to decline. It has to do with the dynamics of costs and benefits.
Unwanted mode changes, one of the costs, have an exponential shape, meaning they start slowly but as the congestion charge increases they increase faster and faster, even though other benefits increase. This explains why net profit, the sum of all costs and benefits, first increases and then decreases, peaking in between.
As we can see in the second graph, the congestion charge that maximizes net profit is $80, which is almost six times higher than the likely congestion charge of $13 to $15. At the optimal toll level of $80, the net societal benefit is nearly $10 billion per year.
The region to the left of the optimal point is the efficient region, where time savings and environmental benefits grow faster than costs. Beyond this point, we enter the region of diminishing returns, where costs – the higher tolls paid by motorists, as well as the drivers’ lost enjoyment of car journeys they forego because of the toll – increase faster than the benefits.
The efficient region of the graph is where the toll offers the best value for money: a $14 congestion toll hits drivers with only 17.5% of the optimal toll cost while extracting 45% of the maximum net social benefits. This is efficiency in action!
Now let’s look at another way to determine the optimal congestion charge. Economic theory suggests that in a situation with a negative externality such as traffic jams, a public policy remedy is to introduce a corrective taxation equal to marginal cost from exteriority.
Let’s unpack this wacky claim: a negative externality occurs when a private party uses a public resource, but the cost of using that resource is borne by the public, while the private user doesn’t pay a penny.
This situation is often called the tragedy of the commons. Traffic congestion is a negative externality where the common public resource – the road – is overused by its free availability and no longer works for the common good. (Besides traffic congestion, another crisis that falls under the common framework tragedy is climate change, as polluters are rarely blamed for releasing CO2 in the atmosphere.) Marginal cost in our context simply means the collective delay costs incurred by all drivers due to an additional vehicle round trip to Manhattan’s central business district, the cordoned off area.
The marginal cost of delay is another output of BTA. Here, I’ve plotted it against the congestion charge:
We can see that the marginal cost of delay at a $10 congestion charge is about $140. This means that when the congestion charge is only $10, the collective delay imposed on all other drivers by an additional car trip in the CBD is financially valued at $140. With a congestion charge of $20, congestion is reduced and, as expected, the marginal cost of delay also decreases (to $100 in this case).
The graph shows that the marginal delay cost curve slopes downward and meets the rising line of the round-trip congestion charge at just over $50. This intersection indicates the toll level for which the marginal cost of delay is equal to the congestion charge. Recall now that the prescriptive public policy remedy for dealing with a negative externality is to charge a corrective tax (our congestion tax!) equal to marginal cost. Thus, the theoretically optimal congestion charge should be around $50 (lower than our previous optimal charge of $80, because this graph only takes into account one externality: delays for drivers).
So let’s review:
- The most likely peak round-trip congestion charge to emerge from the Traffic Mobility Review Board’s review process is between $13 and $15.
- Current optimal the congestion charge, which we calculated in two ways, is between $50 and $80.
- The likely congestion charge between $13 and $15 is an effective range that yields about half the maximum net benefit while charging less than one-fifth of the optimal congestion charge.
While it is great that we get almost half of the maximum societal benefit from the likely congestion charge, it is also unfortunate that we leave more than half of these potential benefits on the table. This represents a huge saving in time, environmental and welfare benefits of $5.4 billion per year. The corresponding toll revenue that the MTA would leave on the table is over $2 billion per year.
I’m not advocating charging congestion charges in the optimal range, but clearly we would do well to stay in the upper end of the likely $13-$15 range. Indeed, looking closely at our second chart, we can see that the difference in net benefit to society from moving from a $13 congestion charge to $15 is nearly half a billion dollars per year, an increase of 11%.
The corresponding increase in toll revenue is $140 million per year, or 12%. These are resources the MTA could use to upgrade subway stations with platform barriers for safety and expedite the deployment of CBTC to increase subway speeds.
Gui Larangeira holds a doctorate. student at the State University of New York at Stony Brook, in the Department of Technology and Society. His research focuses on the modeling of energy and transport.