One Trillion Dollars in Infrastructure

Among President Trump’s pledges is a promise of investing one trillion dollars into America’s infrastructure. Republicans are generally less favorably disposed to large public work projects; this investment is meant to be privately funded, though accomplishing one trillion worth of investment with no public commitment is probably impossible.
In this context, “infrastructure” refers to publicly owned and used transportation assets, such as roads, bridges, airports, rail lines, stations, depots, light rail. A hot topic in infrastructure is the use of public-private partnerships, or PPPs. PPPs are one buzzword being thrown about by a number of pundits without a great deal of dialog about what it entails. Basically, building infrastructure – for example, roads – does not change. The road or asset is designed, specified, and scoped in terms of requirements, land acquired and cleared, prepared, and the road constructed. Of late (last decade or more) there has been a movement away from the classic public acquisition strategy of public transportation officials specifying detailed design of the road and releasing a request for proposals for road builders, as if a road had only one way to be built and DoTs were merely shopping for the assembly. Instead, transportation authorities such as state DoTs now concentrate on forecasting and planning the need, and releasing a design-build request for proposals. An infrastructure construction firm has some ability to configure road path, bridge and embankment design, treatment of problem areas. The movement to design+build helps ease the surprises along the build phase, and provides more alternatives for working around problems. For example, if an important archaeological burial site were discovered while excavating a roadbed, traditionally the construction would shut down until the site could be properly studied and possibly moved, a lengthy delay of months to years. Under design-build, the construction firm may have alternatives for building around the archaeological site from among the original roadbed site selection, and slight changes in design may avoid a lengthy delay.
Under traditional road construction, a state finances road construction with grant money from the federal DoT, or sells bonds to raise a pile of cash to pay for the upfront construction. Federal highway money could be spent a hundred times over; demand far exceeds supply. Bond money must be repaid, from the general state tax fund, from a dedicated tax fund (such as lottery ticket sales, or general sales tax, or taxes on goods such as tobacco.) Offering bonds allows the state to stretch out payments over a much longer period of time. Generally these are tax exempt bonds, and are reasonably attractive assets to pension funds and mutual funds. PPPs offer a new alternative: that a private company (usually a consortium) designs and builds the asset and directly collects the payback. For example, the DoT may provide 10 to 50% of the total construction cost up front with the consortium financing the rest, in exchange for receiving the payback that the state expects, particularly funding from a dedicated tax source. California high speed rail projects are being undertaken with PPPs and a future revenue repayment from California’s carbon tax.

A large number of PPPs for roads are repaid with tolls on the new road. The consortium provides design, build, operation and potentially maintenance of the new road, for some fixed period of time such as 40 years. The integration of design and build with operations and maintenance (DBOM) ensures that the most economically efficient road will be built, as the construction firm is not building with a 3 year time horizon for payoff, but with a 40 year time horizon, identical to the DoT time horizon. Any time-based trade-offs of construction materials against maintainability will be decided with the same time horizon as the road lifetime. Under older models of infrastructure construction, any shortfalls in construction that became apparent over time would be dealt with by the state operator, long after the construction firm had collected its payments for the job. Under a PPP, the construction consortium will share in those shortfall or be wholly subject to them. Thus PPPs have the attractive characteristic of sharing risk across both the DoT and the consortium.
In the case of tolled PPPs, such as a road whose construction cost is repaid with user tolls over the next few decades, toll revenue depends on traffic on the road. Here is another element of shared risk: the manifestation of predicted travel levels. Often after a new tolled road is built, users tell of driving the road in isolation, as few drivers chose to change their driving routes and pay the out of pocket toll. Realized usage may depend on popularity of a new airport at the end of the road, or upon the development surrounding the road as a city spreads. Frequently the DoT will “guarantee” a level of traffic consistent with DoT forecasts. If the usage that the DoT forecasts does not materialize, toll revenues will be lacking, and the consortium would be in danger of failing to repay their own debt on the venture, so the DoT provides a payment if traffic is under their guaranteed level, an insurance against loss. But the payment is generally limited; so the consortium is also bearing risk against traffic not appearing.
Under more sophisticated tolled PPPs, the consortium can end up paying a fee to the DoT if traffic exceeds projections; congestion pricing mechanisms can have complex puts and takes; and public right of way for emergency vehicles and high occupancy vehicles can be part of the agreement.
The bonds that DoTs issue are tax exempt and very desirable to retirement funds, so the cost of money under DoT bonds is very, very low. Commercial bonds are more expensive. Any savings in building infrastructure under PPP does not come from cost of money; it comes from the desire of the consortium to be the operator of a reliable and continuing source of income for 40 years, which may lead companies to cross-subsidize through time. That is, they will bid the construction project with less cost reserve built in if they have a reasonable expectation that any possible construction cost overruns can be made up in future toll revenues.
Consortium construction can be financed through commercial bonds (not tax exempt) and also through equity arrangements. The basic mechanism is the same; a pile of money needs to be generated to pay the upfront construction, and then the money has to be repaid in one form or another. Equity financing is the sale of ownership stock of the ultimate operating company. The operating company will pay dividends on the stock from toll revenue. Equity financing based on excise tax repayment seems less likely, since excise tax revenue rarely grows as fast as toll revenue.
Using PPPs does not require an act of the legislature. For instance California, Texas, and Virginia all use PPPs in road projects, but have no explicit laws authorizing PPPs. However, some budget regulations, such as the City of Arlington, restrict funding redirection from certain revenue inflows – such as excise taxes – to fund things like infrastructure, so in isolated cases, the range of possibilities is not entirely open. Cross-subsidization of high speed rail from a carbon tax required the passage of at least a half dozen state bills. And most of this work falls upon the states, both red and blue, rather than on the administration.