New Jersey toll increases proposed by Corzine were a political decision - REVIEW of T&R reports

Governor Corzine and his administration specified the New Jersey toll increases to be modeled by their traffic and revenue (T&R) consultants in a reversal of the normal relationship. The four reports released by the New Jersey Treasury underline the Corzine administration's primary focus on maximizing revenue for the state from the proposed concession agreement with a new 'public benefit corporation' that would operate the New Jersey Turnpike (NJTP), the Garden State Parkway (GSP), State Route 440 (NJ440), and the Atlantic City Expressway (ACE) for 75 years.

Steer Davies Gleave (SDG) in their four reports for the New Jersey Treasury - this UK based firm was partnered with CRA and EDRG for the studies, and acknowledges help from WSA and Vollmer in their earlier T&R work for the state toll authorities - say that the six toll scenarios they modeled were "defined by the state."

And their reports don't, like many T&R studies, recommend new toll rates.

They do, apparently meticulously, craft reasonable assumptions about likely growth of traffic and value of time, and apply them to modified versions of established traffic assignment models, and generate likely projections for traffic and revenues for each of six toll rate scenarios.

The Corzine administration appears to have chosen to run with the toll scenario that generates the highest toll revenue, Scenario V, which raises inflation-indexed or 'real' tolls:
- 50% in 2010
- 50% in 2014
- 50% in 2018
- 50% in 2022
which we summarize as 'four fifty percenters' or 4x50%.

Another scenario VI has the highest toll increases (2x75%)+(1x50%) and suppresses traffic on the NJTP more, apparently via the shock effect of the 75% increases, not from where toll rates end up (marginally lower).

Elastic elasticity

SDG have an interesting discussion of the central question of likely responses of motorists to toll increases - technically the price elasticity of demand. This is key both because it determines expected revenues and also because it bears on traffic diversion to parallel free routes, although motorists may also react to the higher toll rates by ranging less far afield. Journeys aren't necessarily diverted, simply not made.

Low low tolls now

Currently for most motorists New Jersey toll rates are so low relative to what tolls would be in a real market for roadspace that people don't even bother to know what they are, SDG surveys show.

They write: "The tolls on the New Jersey toll roads are currently low - in comparison with most in the US and certainly in other advanced countries - while the advantages of using the tollroads is high. For most users the level of tolls is well below the indifference price. The toll can for these users be raised very significantly before they will seriously consider using a free alternative." (We've edited their verbose prose a bit TRnews)

There is hilarious survey material presented showing how many people who drive the NJ tollroads know little about the toll rates - apparently because tolls are so trivial in their budgets it isn't worth knowing.

42% of people surveyed think toll rates have gone up in the past two years when there has been no change since 2003! (SDG1 p101)

57% say they don't normally think about the cost of the toll when making a trip. (SDG1 p95). Very few say they worry about the amount they spend on tolls. (SDG1 p97).

This isn't surprising. For most people the tollroads provide value so clearly ahead of the cost of tolls, there is no reason to know what the toll is. You are going to ride the tollroad anyway.

Affluence

By most standards it is rather affluent people who drive the tollroads in New Jersey, the SDG surveys show - motorists on the NJTP are almost exactly equally New Yorkers and New Jerseyans, from the two wealthiest states in the union.

36% of tollroad users have household incomes over $100k/yr and over a half incomes over $75k. (SDG1 T4.5 p88)

So at current low toll rates elasticity is low.

SDG are at pains to point out however that elasticity may be higher at much higher toll rates - the function may be "non-linear" in mathematical jargon.

They did a first phase of modeling with elasticities of about -0.1 meaning traffic would only decline 10% for a doubling of toll rates. (NJTP -0.1, GSP -0.07, and ACE -0.12)

But in Phase 2 they recalculated and used higher elasticities: NJTP, NJ440 and ACE -0.2 to -0.3, GSP -0.07.

They say in reviewing the matter that it made sense to assume there is a higher elasticity in offpeak hours when parallel free routes are running more freely, and that elasticity of demand will be lower, perhaps around -0.1, in peakhours when the alternative routes are very slow.

SDG also discuss how price elasticity of demand may be higher in the case of large shock toll increases, than if the same toll rates are attained in more frequent but smaller toll hikes. Their scenario VI with 75% toll increases but toll rates slightly lower is projected to reduce traffic as compared to the more frequent smaller toll increases but slightly higher tolls of the chosen Scenario V. See bottom right corner of Table 6.3 SDG2 p60 and partially reproduced nearby.

Elasticities clearly vary also by segment of road. Those sections with good parallel free alternatives - like the NJTP in its southern segment with I-295 nearby - clearly have higher price elasticity than others like the north where alternative routes are lacking.

Despite this interesting discussion it isn't clear exactly how SDG applied elasticities in their modeling.

We have traffic projections in the chosen Scenario V (the 4x50%) that substantially reduce traffic on the tollroads and even more substantially reduce traffic as compared to the growth that would occur with toll rates kept at their present real level - that is increased each year in $-terms in step with inflation - Scenario I.

For the NJTP traffic declines by 39% 2008 to 2022 at which point constant real tolls would have seen traffic increase by 36%. So traffic is down almost 55% on what it would otherwise have been. From 2022 onward the gap closes slowly because under Scenario V there are no further toll increases but growing percapita income and value of time saved.

The reductions in traffic are less on the GSP but more on the ACE.

As long as elasticity of demand is less than -1.0 (such as -0.5 or -0.8) the higher toll rates and reduced traffic of Scenario V still make for higher revenues. In 2022 the NJTP revenues are $1882m compared to $770m of Scenario I's merely indexed toll rates. (All tolls are reported in 2006 prices so all scenarios get indexed.)

Diversion or suppression or here does the traffic go?

A glaring deficiency in the work done by SDG is the lack of any discussion of reduced traffic:

- how much is diverted onto parallel free roads

- how much changes mode (to rail)

- how much is trips just not made, with people shopping or working closer to home

The 'diversion' issue looms especially on the GSP which already is a mess of untolled segments and tolled one-way, untolled the other. This is a minor problem when tolls are so low - as now - that a substantial portion of motorists don't even bother to inform themselves what the toll is.

But throw in several 50 percenter increases and that will change. Shunpiking will arise with a vengeance.

New tolling regime?

Is the public concessionaire company going to be allowed to fill in the gaps in the tolling format of the GSP or will the present tolled and untolled sections be frozen for the life of the concession?

Cash toll collection at the present low toll rates argues in favor of the open-way tolling and some ramps left untolled. The costs of toll collection don't warrant a socalled closed toll system. But in five years time many tollroads in America will have gone all electronic.

Also higher toll rates on the GSP will take tolls beyond the levels that can be conveniently paid in coins, so coin machines will be scrapped. Would expensive bill accepting machines or toll booths be installed instead. Unlikely. Logic will point to scrapping coin machines and going to all electronic tolling - transponders and cameras for billing by mail.

With all electronic tolling it becomes economical to extend the gantries over both directions of traffic and to collect two-way tolls again. Countering diversion will be another justification for that.

However there will be mini-constituencies of motorists who stand to lose from more comprehensive tolling, so it will be controversial. It remains to be defined how the proposed concession will control or allow such changes in where and how tolls are collected.

Variable pricing

SDG touch on variable toll rates. The NJTP already has a 25% discount with transponder (E-ZPass) tolls offpeak hours, probably not large enough, they say to change the distribution of trips by time, though it may encourage a few motorists to travel early. The GSP once had a small discount but it was dropped in 2001 in an effort by the state to get a little extra revenue without announcing a formal toll increase.

Given the lower elasticity of demand in peak hours, and the consequent lesser diversion, more highly differentiated toll rates by time of day would make sense in both boosting revenue and reducing diversion. The concession will have to spell out whether and on what terms the concessionaire can have differentiated tolls.

Early revenues are much more important than later revenues in any case because of present value, but they are especially central to the Corzine scheme because all the real toll increases are proposed in the first 12 years. The SDG reports don't sum them but we've done it for 2022 in the nearby table. The 4x50% scheme (4TRs V) is projected to generate $2970m or 3.65 times the $814m being garnered now. At constant real toll rates (4TRs I) revenue will grow from $814m to 1.44 times that to $1173m. (Table between the map and the governor above)

Toll rates under Scenario V grow 5.06x but traffic is lost to produce the lesser 3.65x increase in toll revenues. SDG provide projections year by year out a hundred years, but the 2022 numbers probably are the most relevant to the value of the concession.

Download the four SDG reports under "Hot topics" in the middle column of the New Jersey Treasury website:

http://www.state.nj.us/treasury/

Concession contract the key

Many of the details of the Corzine tollroads scheme have yet to be finalized, it appears. Most notably there is no draft concession contract. The concession is key to the whole scheme. By giving the concessionaire a contractual right to 4x50%+inflation toll increases the project is protected from political backsliding, and is able to raise the $30+ billion.

A plan to have the existing Turnpike Authority and SJTA (ACE owner) increase tolls by the same amounts - even if endorsed unanimously by the legislature - would carry little weight with bond investors because a subsequent governor and subsequent legislators would be tempted to jettison the plan and impose a toll freeze. But armed with a 75 year concession contract the concessionaire company will have an irrevocable right to levy tolls up to the high toll rates.

The US Constitution has been successfully invoked to protect contracts against state efforts to break them, whether by state law or even state constitutional amendment.

See the bolded item of Section 10 of the US Constitution. which also interestingly precludes Gov Corzine from minting his own money or offering the head of the concession corporation a Lordship or Barony: "Section 10. No state shall enter into any treaty, alliance, or confederation; grant letters of marque and reprisal; coin money; emit bills of credit; make anything but gold and silver coin a tender in payment of debts; pass any bill of attainder, ex post facto law, or law impairing the obligation of contracts, or grant any title of nobility."

Also needed to reassure bond investors will be provisions that give the board of directors independence of the state government of the day. That's difficult to guarantee when it's the government that appoints them. And there's always the conundrum of "independent" government entities: who are they answerable and accountable to?

Private concessionaire are accountable to their shareholders. A normal public toll operation will be accountable to the state government of the day. It won't be clear who an independent public concessionaire will be accountable to, so the criticism will inevitably arise: "They are a power unto themselves, a self-serving, self-perpetuating empire" etc.

Still the concession contract will need to allow the concessionaire to commit itself to maintaining and improving the roads according to some schedule, and to rank servicing of debt next - well ahead of political considerations.

Brad Abelow the governor's chief of staff said last week: "For us to engage in this transaction requires us to be able to say to the financial market...that this company is truly independent, the Public Benefit Company will be truly independent of the State."

Of course the New Jersey Turnpike Authority and other state toll agencies were originally constituted independent of the state to raise their original capital. That's why they weren't mere divisions of the state DOT. But over the years they have been progressively brought under political control.

You have to be somewhat skeptical about this aspect of the Corzine scheme.

There's a lot we don't know but we do know:

- the state's new concessionaire is provisionally named the Public Benefit Corporation or Company

- the board of directors of the PBC will be a "broadly representative of the folks that care about (this)" (Brad Abelow)

- PBC expected to raise $32.5b to $37.3b in debt capital (it won't issue stock since the state will own it 100%)

- PBC will be awarded without competitive bids a 75 year concession to operate and toll the New Jersey Turnpike, the Garden State Parkway, a short connecting section of NJ440 and the Atlantic City Expressway

- unlike the New Jersey Turnpike Authority the Governor will have no "minutes veto" powers to override the PBC directors

- no directors of the PBC will be appointed by the Governor (but who will appoint them and for how many years?)

- the PBC will have the rights to the service area concessions, naming rights and other sources of non-toll revenues

- if the PBC has surplus revenues they can be used to keep tolls lower than otherwise or to support the state's transportation finance needs

Brad Abelow the governor's chief of staff has said the PBC will have a "large sized financial reserve set up in the initial funding" to protect against any shortfall in revenues. This need is one of the weaknesses of all-debt state entities as compared with investor owned companies as concessionaires.


COMMENT: The Corzine administration deserves credit for moving to thaw out New Jersey toll rates. They have been frozen so long they no longer function as prices should. They fail to provide signals to motorists about the costs they impose by driving. And they fail to generate the return on capital that their valuable mobility services would in the absence of political intervention.

Tolls should be roughly adjusted to some measure of inflation. The notion that somehow toll operators alone in the economy can forgo inflation adjustment is preposterous. Politicians who have frozen toll rates in nominal terms, constantly eroding their real value are irresponsible panderers.

Corzine deserves respect for demanding a break with that bad practice.

Toll rates in New Jersey should be somewhere between three and four times their current rate - 7c/km (12c/mile) to 15c/km (24c/mile) for cars and a multiple of that for trucks.

However the 4x50% increase or a five-fold increase in real terms 2010 to 2022 is probably excessive.

If the elasticity of demand of shock toll increases is greater, as SDG say, then maybe they would be better done as annual increases of 10+% than as 50 percenters. (see nearby the comparison of Scenario VII, which we made up)

If the five-fold real increase in toll rates is overdoing it, it needs to be said that the Corzine administration has proposed these toll rates as caps - maximum toll levels allowed to the concessionaire. Reporters have glossed over that distinction and called them proposed toll rates. So the PBC concessionaire could keep toll rates lower. There is downside flexibility.

You have to be skeptical about the PBC form proposed. Saying it is intended to be businesslike and independent cannot make it so. It will lack any accountability to shareholders. It will also lack the incentive to efficiency of a private entity that gains a concession in a competitive bid and is expected to perform in maximizing revenue and minimizing costs.

Private firms have the advantage for society of being subject to takeover, and having shareholders with the power to throw the managers out when they aren't performing. Conversely private sector managers get rewarded handsomely when they do perform. No one has worked out a way of making that happen in the public sector.

Gov Corzine has embraced the principle of a private toll concession, and proposes to implement some of its features. But he has copped out of the difficult politics of doing the real thing. We'll see whether this hybrid will fly.

CORRECTION NEEDED: A correction is needed to SDG II p4. The NJ Turnpike does NOT have an interchange Exit 3 with ACE as claimed by SDG. Exit 3 of the Turnpike is an interchange with a signalized arterial, NJ168, over a mile away from where the ACE and the Turnpike cross without ramps. To get from the Turnpike's Exit 3 onto the ACE at its Exit 12 with NJ41 is an easy-to-get-lost 10 minute journey of 4km (2.5 miles) on surface streets. (see our satellite map with annotations nearby) The non-existent interchange SDG describe between ACE and the Turnpike is a badly needed improvement.

TOLLROADSnews 2008-01-21