TxDOT may subsidize and ensure profits for private toll roads

[wxfree]

I'm not sure I fully understand the proposed changes, but based on my own reading and commentary by others, it looks like the Texas Transportation Commission has proposed rule changes that may be illegal, and may be politically unpopular.  The changes would allow the department to enter into a pass-through toll agreement with a private developer and make payments for expenses "including financing costs, and to pay a return on any private investment."

Pass-through toll agreements are, in my view, a "Pee-Wee's Playhouse" method of funding roads.  It's an "innovative financing" method (financial innovation is part of what gave us the financial crisis).  The city of Weatherford paid for improvements to certain highways there, and in exchange TxDOT will make agreed-upon payments based on road usage, basically making them into toll roads for which TxDOT, rather than drivers, will pay the tolls to the city to recover the cost.  It's essentially another form of debt.

I find this agreement between TxDOT and a city questionable, but they're wanting to make it worse.  Under the proposed rule change, TxDOT would be able to enter into a pass-through agreement with a private developer with the following provision

Payments may be made in connection with a highway project financed under an agreement procured under §5.61 of this subchapter in order to reimburse the private entity's project-related costs, including financing costs, and to pay a return on any private investment.

Pass-through tolls can be placed on regular toll roads.  This provision appears to allow TxDOT to subsidize and guarantee profits on private toll roads.  The reason is because some roads may not be fully self-supporting as toll roads but still present a public good.  In order to get the road built, TxDOT may have to assure a return on investment.

On the one hand, it may be unseemly to guarantee profits to private road companies with tax money.  I, myself, am not a fan of any private profit from toll roads that get any government help, such as eminent domain.  On the other hand, the state may benefit from the roads being built that might not otherwise get built, and the payments from TxDOT would be far less than the actual cost of the road.

This could, perhaps, be put in the General Highway Talk forum, because it's about the philosophy of paying for roads moreso than this proposal by TxDOT.  I don't see roads as free-market products that are worthwhile only if enough people are willing to pay to use them.  I see roads as a public good that is rightly subsidized and produces an enormous return on investment by essentially enabling the entire economy to exist.  I see tolling roads as an option of last resort.  I think that tolls should be just high enough to pay for the road, and not used as a source of revenue for other things.  I don't think private companies should profit from operating a toll road unless they built it without any state help, especially eminent domain, but I don't refuse to drive on such roads because the few dollars they do or don't get from me aren't enough to make any difference.

The link to the proposed rule changes is below.  The changes may be illegal, and may be unpopular.  Pragmatically, I can see the value in ensuring profit for a toll road, effectively getting a road built for much less public investment than the cost of the road.  What are your thoughts?

http://ftp.dot.state.tx.us/pub/txdot-info/adm/2013/documents/minute_orders/may30/9b2.pdf

[MaxConcrete]

As you mention, this is just a mechanism for debt-financed highway construction but the debt does not show up on the financial books of the State of Texas.

I know TxDOT is carrying a large amount of debt, and may not be able to carry any more. Or, any further debt could cause a lowering of the state's credit rating. So this mechanism is way to take on more debt.

Nevada is doing the same thing with its expansion of Interstate 15, headline "Nevada To Try P3 for Massive Neon Highway Project".

In my opinion, I would rather see these projects paid for using enhanced existing revenue streams. But "pass-through tolls" are preferable to actual user-paid tolls, because ion Texas once a road is tolled it will never be untolled.

[J N Winkler]

My thoughts:

*  Pass-through tolls as outlined in this rule change are essentially a variation on the shadow tolling concept, only in this case it does not seem to exclude the possibility of the toll concessionaire double-dipping--in other words, collecting a toll both from the user and from TxDOT.  In a double-dipping scenario, an offer on TxDOT's part to pay shadow tolls could work as a deal-closer.

*  To an extent, pure shadow tolling (no direct tolling of users) tends to shift opportunities for moral hazard away from the concessionaire and toward the owning agency.  In Britain the Highways Agency (an executive agency responsible for trunk roads in England) used shadow tolls to build near Leeds a length of the M1 motorway north of the M62 known as the "M1-A1 Link."  Signing, however, encouraged--and continues to encourage--motorists to use the longer route via the M62 and A1(M), to keep the Highways Agency's shadow tolling bill down.  This is arguably an instance of moral hazard on the owning agency's part.  This is a contrast to the usual position in Texas, where a concessionaire gets the right to charge tolls through a CDA or similar instrument and then cranks the tolls high to keep traffic low and limit maintenance costs, so the public gets little benefit from the brand-new road.  Moreover, to the extent that heavy traffic exposes unforeseen shortcomings in design and construction that disadvantage the concessionaire, these are more likely to surface under shadow tolling arrangements where the up-front cost to drivers is far less and traffic is therefore heavier.

*  Shadow tolling locks up TxDOT's cashflow a little at a time.  The effect becomes progressively worse, depending on what arrangements are made for index-linking the per-vehicle shadow toll and how these relate to increases in TxDOT's revenue through fuel taxes and other sources.  Money that is committed to paying shadow tolls is unavailable for maintenance and construction.

*  If traffic on a shadow-tolled road is much greater than forecast, then it blows a hole through TxDOT's budget.

The only current problem I see that could be attacked through a shift to shadow tolling is severe underuse of newly built toll freeways.  It removes the deadweight cost of collecting toll directly from the users, but at the risk of creating severe fiscal headaches for the owning agency.  TxDOT is still much better off resorting to an increase in the fuel tax to finance unmet needs.  The problem is that the Texas legislature will never agree to such an increase as long as it continues to see the highway finance problem in false terms.

On the whole I am inclined to see a shift toward shadow tolling as a step--but only a step--in the right direction, rather than as yet another invitation to the private sector to help itself to gas tax revenues and tolls.  But theory is one thing, and practice quite another.  Shadow tolling could wind up being even worse for Texans than the current generation of tolling agreements if TxDOT acquiesces in one-sided agreements (e.g., ones allowing double-dipping) or sets itself up to be milked by higher-than-expected traffic.

[Chris]

Shadow tolling is often seen as unfavorable to the government. Often traffic volumes turn out to be higher than forecasted, resulting in a higher cost for TxDOT and a lower remaining budget for other tax-funded roads.

The Netherlands built a shadow toll tunnel, and the total cost over 30 years turned out to be three times higher than if it was constructed with government funding. Portugal recently eliminated pretty much all shadow toll roads as an austerity measure. Regular license plate tolling was introduced, resulting in traffic drops of 40 - 60% in many cases, and a shift of traffic to non-tolled roads, increasing maintenance cost for those. (not to mention the adverse result on traffic safety, because limited-access highways are much safer than rural two-lane roads).

You don't see many new shadow tolls in Europe anymore. They usually go for a P3 project based on availability payments. Germany has a few situations with a hybrid system where the operator gets the truck toll revenue (passenger cars are not tolled in Germany). That turned out to be a gold mine for the operator and a loss for the tax payer.

[Alps]

This is the first I've heard about shadow tolls, and after starting to learn about it, I was dumfounded (yes, that's the spelling) by Wiki's inclusion of the Dulles Greenway as a shadow tolled route. It's not, right? It has tolls on it.

[NE2]

I just removed it and the 495 express lanes. The source cited didn't support it (and if it had, it would have been wrong).

[wxfree]

What's being presented here is a change that would allow TxDOT to ensure profits for private toll roads.  The reason is that there may be value in getting a road (even a toll road) built that might not get built without subsidies to cover losses.

Current rules prohibit paying a return on investment to a private entity

Under the existing program, the department can only reimburse a private entity what it would cost the department to construct the project if a pass-through toll agreement was not used.

The proposed rule change would allow the department to pay "a reasonable rate of return on investment."

There are critical highway improvement projects the department cannot currently construct because there is insufficient funding or other financing methods available to do so. There exists the potential to use a pass-through toll agreement to obtain private financing of these priority projects. In order to obtain private financing, the department will need to reimburse the private entity's financing costs and to pay a reasonable return on investment, which is prohibited under the existing program.

This could apply to a company paying upfront to build a road and paying for ongoing maintenance, and then getting payments over years from TxDOT at a level that would pay a return on that investment.  Even if a maximum payment level is defined, this seems hokey to me, like trying to get into debt and hide it.

Regular (drivers pay) toll roads can also be put under pass-through agreements.  Pass-through payments can be adjusted based on whether the road is tolled.  This proposed change presents the potential for a company to pay to build a toll road and use a pass-through agreement to receive payments from TxDOT in addition to toll revenue, generating a "return on investment," profit.  The southern extension of State Highway 130 is losing a lot of money, and the concession company may go bankrupt.  If they had a pass-through agreement allowing TxDOT to pay a rate of return, that could results in payments (possibly in limited amounts) to cover the losses and generate a profit.

What's most contentious about this is the possibility of ensuring profits for private toll roads at the expense of the highway fund.  There may be public value in getting a road built, but if you're already going to put tolls on a road, and guarantee the debt with the highway fund (like the CTTS), why involve private companies with the added expense of a subsidized profit?  Why would TxDOT not build the road itself, collect the tolls, cover the losses, and keep any potential profit?  This would add more debt, but it's a more honest way to get into debt, it brings a dedicated toll revenue stream, and doesn't give away potential profit.

[NE2]

Why would TxDOT not build the road itself, collect the tolls, cover the losses, and keep any potential profit?

Because gubmint profit is bad and corporate profit is good.

[wxfree]

This is the first I've heard about shadow tolls, and after starting to learn about it, I was dumfounded (yes, that's the spelling) by Wiki's inclusion of the Dulles Greenway as a shadow tolled route. It's not, right? It has tolls on it.

I don't know about the Dulles Greenway, or shadow tolling in Virginia, but in Texas a toll road can also have pass-through tolls.  Texas Administrative Code Title 43 (TxDOT rules) Chapter 5, section 5.58 (a)(13) requires that the proposer indicate whether a project is to be tolled.  http://info.sos.state.tx.us/pls/pub/readtac$ext.TacPage?sl=R&app=9&p_dir=&p_rloc=&p_tloc=&p_ploc=&pg=1&p_tac=&ti=43&pt=1&ch=5&rl=53

I've read that whole chapter and still don't fully understand how the whole thing works.  I need to dig up an actual agreement and see if I can figure it out.  I can't fully interpret the proposed rule change without more knowledge of the existing system, but multiple news articles seem to interpret the changes the same way I do.

[wxfree]

Because gubmint profit is bad...

I mostly agree with that.  If your toll road is generating more money than is needed to pay for it, to me that means you're charging your drivers too much.  Of course, if you're going to have a policy of overcharging the public, the benefit should go to public, not private, interests.

[J N Winkler]

What's most contentious about this is the possibility of ensuring profits for private toll roads at the expense of the highway fund.  There may be public value in getting a road built, but if you're already going to put tolls on a road, and guarantee the debt with the highway fund (like the CTTS), why involve private companies with the added expense of a subsidized profit?  Why would TxDOT not build the road itself, collect the tolls, cover the losses, and keep any potential profit?  This would add more debt, but it's a more honest way to get into debt, it brings a dedicated toll revenue stream, and doesn't give away potential profit.

I agree with your point about maintaining the transparency of TxDOT's debt obligations, and this change should certainly not be adopted with the intention of keeping added debt off TxDOT's balance sheet.  But if a debt vehicle of some sort is going to be used to build a proposed new road, whether TxDOT itself issues the debt or asks a private company to borrow the money on its behalf in anticipation of future repayment, then I don't see anything fundamentally wrong with a profit guarantee so long as it is relatively low, is capped, requires the company to carry out its business in a reasonably efficient way, does not oblige TxDOT to do anything if the company is already making reasonable profits, and is part of a framework agreement between TxDOT and the company that guarantees value for money.  The reason the profit guarantee should be low is that by definition it implies TxDOT, and not the company, is assuming the investment risk of the project.

It is important to realize that with any of the current and proposed financing methods available to TxDOT, there are opportunities for the private sector to turn a profit.  If TxDOT designs the complete project and puts it out to bid, for example, then the consultant (if TxDOT uses one) responds to the design services solicitation in the expectation of making a profit, and the contractor places a bid with a like expectation of making a profit.

It is certainly true that TxDOT does not, as a matter of course, guarantee contractor profit through the design-bid-build process, but if TxDOT is anything like Kansas DOT, it guarantees profit for the consultant through a cost-plus payment system:  actual cost of services rendered plus overhead at a set percentage plus profit.  (Fixing overhead at a set percentage gives the design consultant an incentive to run its office efficiently while retaining enough capacity to be competitive for soliciting agencies' design work.)  And even in the absence of a profit guarantee for contractors, most state DOTs try to structure their construction contracts so that there is a reasonable level of competition among bidders, but with enough clarity about the risks to allow the winning bidder for a given project to build it with a reasonable chance of making a profit.

No state DOT wants to have so many business partners (such as design consultants and construction contractors) drop out that the few who are left can charge monopoly rents for their services.  Profit by definition implies a departure from perfect competition, but it is often the price that has to be paid to ensure that the capacity is there when the agency needs it.

The more serious problem with new freeways in Texas is the persistent attempt to charge toll on them.  It makes no sense to build to accommodate heavy traffic demand, scare off nearly all of that anticipated usage by charging excessive tolls, and add insult to injury by paying heavy collection costs.  If there is a new method of financing new roads that promises to ensure they get built at a reasonable final cost to the taxpayer (whether private companies make a profit or not) while eliminating the deadweight cost of toll collection, then the potential benefits both to taxpayers and motorists demand that it be given careful consideration.

[wxfree]

What's most contentious about this is the possibility of ensuring profits for private toll roads at the expense of the highway fund.  There may be public value in getting a road built, but if you're already going to put tolls on a road, and guarantee the debt with the highway fund (like the CTTS), why involve private companies with the added expense of a subsidized profit?  Why would TxDOT not build the road itself, collect the tolls, cover the losses, and keep any potential profit?  This would add more debt, but it's a more honest way to get into debt, it brings a dedicated toll revenue stream, and doesn't give away potential profit.

I agree with your point about maintaining the transparency of TxDOT's debt obligations, and this change should certainly not be adopted with the intention of keeping added debt off TxDOT's balance sheet.  But if a debt vehicle of some sort is going to be used to build a proposed new road, whether TxDOT itself issues the debt or asks a private company to borrow the money on its behalf in anticipation of future repayment, then I don't see anything fundamentally wrong with a profit guarantee so long as it is relatively low, is capped, requires the company to carry out its business in a reasonably efficient way, does not oblige TxDOT to do anything if the company is already making reasonable profits, and is part of a framework agreement between TxDOT and the company that guarantees value for money.  The reason the profit guarantee should be low is that by definition it implies TxDOT, and not the company, is assuming the investment risk of the project.

It is important to realize that with any of the current and proposed financing methods available to TxDOT, there are opportunities for the private sector to turn a profit.  If TxDOT designs the complete project and puts it out to bid, for example, then the consultant (if TxDOT uses one) responds to the design services solicitation in the expectation of making a profit, and the contractor places a bid with a like expectation of making a profit.

It is certainly true that TxDOT does not, as a matter of course, guarantee contractor profit through the design-bid-build process, but if TxDOT is anything like Kansas DOT, it guarantees profit for the consultant through a cost-plus payment system:  actual cost of services rendered plus overhead at a set percentage plus profit.  (Fixing overhead at a set percentage gives the design consultant an incentive to run its office efficiently while retaining enough capacity to be competitive for soliciting agencies' design work.)  And even in the absence of a profit guarantee for contractors, most state DOTs try to structure their construction contracts so that there is a reasonable level of competition among bidders, but with enough clarity about the risks to allow the winning bidder for a given project to build it with a reasonable chance of making a profit.

No state DOT wants to have so many business partners (such as design consultants and construction contractors) drop out that the few who are left can charge monopoly rents for their services.  Profit by definition implies a departure from perfect competition, but it is often the price that has to be paid to ensure that the capacity is there when the agency needs it.

The more serious problem with new freeways in Texas is the persistent attempt to charge toll on them.  It makes no sense to build to accommodate heavy traffic demand, scare off nearly all of that anticipated usage by charging excessive tolls, and add insult to injury by paying heavy collection costs.  If there is a new method of financing new roads that promises to ensure they get built at a reasonable final cost to the taxpayer (whether private companies make a profit or not) while eliminating the deadweight cost of toll collection, then the potential benefits both to taxpayers and motorists demand that it be given careful consideration.

Contractors usually make a profit when building a road under a normal contract.  I'm not opposed to profit.  I have reservations about putting the risk on the taxpayers and giving not just potential profit, but guaranteed profit to a private company.  But my reservations are not beyond convincing.

If it's already determined that a new road will have tolls (which in Texas means almost any new road), I can understand if TxDOT considers the road to be a public value and being willing to subsidize it if toll revenue is low.  I'm not in favor of it, but I can see the logic behind the thinking.  Even in the worst case scenario, the total paid out by TxDOT would be less than if they'd paid for the road themselves.

Another issue is that it may be illegal.  TxDOT can't guarantee other entities' debt, even public entities.  They had to do financial gymnastics to put together a deal to effectively, but not technically, back up NTTA's debt on their two newest roads.  I'm not sure what their reasoning is on how this arrangement would be legal, but they've shown themselves to be adept at getting around the law.

This issue is very interesting to me, involving highways, law, and complicated arrangements.  The attorney general could weigh in on the matter.  In Texas an official opinion issued by the attorney general has the effect of law unless overturned by a court or legislative action.  I look forward to seeing how it turns out.

[J N Winkler]

Contractors usually make a profit when building a road under a normal contract.  I'm not opposed to profit.  I have reservations about putting the risk on the taxpayers and giving not just potential profit, but guaranteed profit to a private company.  But my reservations are not beyond convincing.

It is for precisely this reason that I suggest a profit guarantee should be conditional.  I don't envisage the taxpayers' interest being well served if it is easier to make a profit as a proposer of a pass-through project than as a design consultant or a contractor.

If it's already determined that a new road will have tolls (which in Texas means almost any new road), I can understand if TxDOT considers the road to be a public value and being willing to subsidize it if toll revenue is low.  I'm not in favor of it, but I can see the logic behind the thinking.

I am not in favor of this scenario at all.  It rewards pass-through proposers for a problem that is of their own making:  i.e., building the road and then charging tolls so high almost nobody wants to drive it.  If a pass-through PPP has to be used, the taxpayers' interest is better served by either shadow tolling with a capping mechanism to ensure that TxDOT is not soaked if traffic turns out higher than expected, or by a PFI involving the availability payments Chris describes.  (In addition to highways, this has been used for NHS hospitals in Britain, the Athens airport, etc.)

Even in the worst case scenario, the total paid out by TxDOT would be less than if they'd paid for the road themselves.

Looking at the cashflows internal to TxDOT gives a false picture of the problem.  TxDOT is less out of pocket, yes, but perhaps not much less if shadow tolling is used without any element of direct tolling and traffic is much higher than forecast.  If direct tolling is used, the subsidy to the pass-through proposer is structured as a simple profit guarantee with no relation to traffic, and the proposer then elects to hike tolls high enough to minimize usage, then TxDOT is stuck paying nearly the entire capital cost of a new road which does not benefit the community at large since almost nobody uses it.

In short, I don't think this rule change benefits the community at large if it is treated simply as a bolt-on to the existing machinery for procuring private-sector toll roads through CDAs and pass-through agreements.  On the other hand, it is appropriate as part of a completely different framework where a company borrows on behalf of TxDOT (or another agency) to build an untolled road, in return for shadow tolls or availability payments spread over succeeding years.  That appears to be the context in which this idea has developed in Texas--I assume the Weatherford city streets for which TxDOT is now paying shadow tolls are untolled?

Another issue is that it may be illegal.  TxDOT can't guarantee other entities' debt, even public entities.  They had to do financial gymnastics to put together a deal to effectively, but not technically, back up NTTA's debt on their two newest roads.  I'm not sure what their reasoning is on how this arrangement would be legal, but they've shown themselves to be adept at getting around the law.

I think that in order for a ban on debt guarantees to be held against TxDOT, it would have to be shown that the profit guarantee is substantively a debt guarantee.  That could be very difficult since a private entity wishing to raise capital can sell equity shares as well as issue debt.  And since you yourself pointed out a while ago that the trend is now against TxDOT having underlying ownership of new toll roads, it could be argued that it is now the norm for a private entity to leverage its (future) ownership of the underlying asset to serve as collateral when issuing debt or as part of its market capitalization when selling shares.

Even so, there is abundant precedent for controversy over debt guarantees.  The early Italian autostrade were built on the basis of state guarantee of the loans, which led to difficult questions being asked in the early 1930's when traffic turned out to be considerably less than expected.  Refusal to grant loan guarantees, on the grounds of high risk, also led to several British motorway proposals being scuttled in the mid-1920's.  This experience, again, suggests that it would be highly inadvisable for TxDOT to offer a profit guarantee except in cases where project scope is well-defined, risk is tightly fenced, and the pass-through proposer is merely paid for availability.

This issue is very interesting to me, involving highways, law, and complicated arrangements.  The attorney general could weigh in on the matter.  In Texas an official opinion issued by the attorney general has the effect of law unless overturned by a court or legislative action.  I look forward to seeing how it turns out.

It will be interesting indeed--I just hope the outcome is consistent with maximization of consumer's surplus, which is my preferred criterion for running the highway system in general.

[lordsutch]

*  To an extent, pure shadow tolling (no direct tolling of users) tends to shift opportunities for moral hazard away from the concessionaire and toward the owning agency.  In Britain the Highways Agency (an executive agency responsible for trunk roads in England) used shadow tolls to build near Leeds a length of the M1 motorway north of the M62 known as the "M1-A1 Link."  Signing, however, encouraged--and continues to encourage--motorists to use the longer route via the M62 and A1(M), to keep the Highways Agency's shadow tolling bill down.  This is arguably an instance of moral hazard on the owning agency's part.

Unless I'm missing something, the advance M1 at M62 signage seems to favor staying on the M1 to get to the A1, at least according to GSV. http://goo.gl/maps/1Zmz3

[J N Winkler]

Unless I'm missing something, the advance M1 at M62 signage seems to favor staying on the M1 to get to the A1, at least according to GSV. http://goo.gl/maps/1Zmz3

It looks like the position has changed fairly recently (the StreetView imagery dates from November 2012), rendering my information out of date.  I have been told that it was a deliberate decision, made around the time the M1-A1 Link first opened to traffic in 1999, not to provide signing that would encourage its use, thus leaving motorists to discover it on their own.