Texas and other states are increasingly using public‐​private partnerships (P3s) to fund transportation projects. However, not everyone understands that there are two different kinds of P3s. They are so different from one another that I think of one as the “good P3” and the other as the “evil P3.”

Under the good P3, which is technically known as demand risk, the public invites a private party to build a facility and to charge for the use of that facility. The collected fees collected are used to pay the cost and perhaps leave some left over for profits. At some point, often after 30 or 40 years, ownership of the facility is transferred to the public.

Under a demand‐​risk P3, nearly all of the costs, except perhaps the cost of the land the facility is on, are paid for by the private party, who accepts all of the risk that users might not be willing to pay enough to cover those costs. If they don’t, the investors lose their money, but the facility remains for someone else to operate. In short, there is little risk to taxpayers.

Demand‐​risk P3s can be found in the form of toll roads in at least ten states including California, Florida, and Texas. In Texas, for example, demand‐​risk P3s have built new roads in Dallas, Harris, and Tarrant, and Travis counties. Some have been more successfulthan others, but all were built at little or no risk to taxpayers.

If you are used to untolled roads, you might not like paying tolls, but toll roads have distinct advantages over other roads. For one thing, they are usually fair: in most states the tolls people pay go for the roads they drive on and not for something else. In Texas, for example, more than half of all gasoline taxes are spent on things other than roads, but 100 percent of tolls go to the roads they support. Toll roads also tend to be the best maintained roads because their managers know if they allow them to deteriorate, people will be less inclined to pay to use them.

In the evil P3, technically known as availability payment, the government contracts with a private operator to build and operate a facility and agrees to pay the private party, out of tax dollars, whether anyone uses that facility or not. Where most of the risk of a demand‐​risk P3 is borne by the private party, the risk of an availability‐​payment P3 is almost entirely borne by the taxpayers.

Availability‐​payment P3s have built a few roads but are always used for P3 transit projects. That’s because, for some reason, no one expects transit to earn a profit. Where the private parties in a demand‐​risk P3 have an incentive to build and operate as efficiently as possible, the incentives are far weaker for availability‐​payment P3s, which are just as subject to bloated budgets and cost‐​overruns as projects built by the government.

The main advantage of an availability‐​payment P3 is not that it is efficient but that it allows government agencies to evade their legal debt limits. For example, in 2004 Denver voters agreed to increase sales taxes to build several new rail transit lines. Almost as soon as the votes were counted, the projected costs of those light‐​rail lines almost doubled. The transit agency calculated it could still build most of the lines by borrowing against the sales tax revenues for 40 or more years rather than just 30. But it had a problem: the law limited the amount of money the agency could borrow.

The agency decided to build two of the rail lines using an availability‐​payment P3. The private parties would borrow more than a billion dollars and the agency would agree to pay the private parties an annual fee that just happened to be enough to cover the private parties’ debt and the operating cost of the rail line. Since the debt was on the books of the private party, and not the agency, it didn’t count against the agency’s debt limit.

Call me old fashioned, but I happen to think that transportation facilities should pay for themselves out of user fees. Not only will this save taxpayers money, it creates an incentive to build cost‐​effective facilities as efficiently as possible.

Once we agree that it is okay to fund transportation projects that lose money, it is an amazingly small step to go from using tax dollars to buy $500,000 buses to using tax dollars to build $5 billion light‐​rail lines. The absurdity of taking this step is that buses can move more people, faster, more safely, and to more destinations than light rail, for a lot less money. If we insisted that both highways and transit pay for themselves, no transit agency would ever propose a new rail line.

Demand‐​risk P3s are the good P3s because they are funded out of user fees at little or no risk to taxpayers. Availability‐​payment P3s are the evil P3s because they often support bloated projects that are funded out of tax dollars at little or no risk to the private parties. Transportation users and policymakers should keep this distinction in mind when reviewing state and local proposals for public‐​private partnerships.