No, you seem to totally misunderstand the point of the exercise. If we take the cost of constructing a mile of freeway as a constant per mile, the total length is irrelevant. The cars passing over any given mile upon it must pay for that mile. Saying a 30 mile freeway makes more money than a 5 mile one is besides the point. It may make six times as much, but obviously it also costs six times as long. Unit costs and revenues are identical.
Multiply that by 365 and we have about $126 million a year, or about $3.15 million per mile per year on a 40 mile highway. Subtract the maintanence from that and you get revenues of about $3 million a mile, which recovers construction costs @ $20 million a mile (current I-10 estimates say $800 million for the expansion's construction costs over 40 miles = $20 million a mile) in about 7 or 8 years.
Is the construction being financed with revenue bonds (normal method of construction of public works)? If so, you need to add the interest charges. Presumably the property underneath the road was once privately owned. You need to add on the foregone property tax revenue on that land. You need to add in the cost of policing the road and responding to accidents and incidents. Lastly, is the construction estimate in current year dollars, or year of expenditure?
I will grant the basic point - freeways in Houston are very heavily used but cheap to construct and therefore generate an atypical amount of money versus what occurs elsewhere in the country.
...which is an erronious and distorted methodology in itself. Your formula is flawed, plain and simple. Highway construction costs as a whole are constant. The number of trips AND the length of those trips are the variables. Since a car travelling 20 miles uses more gasoline and thus pays more taxes than a car travelling 1 mile, trip length MUST be considered in ANY attempt to estimate gasoline tax revenue for a given road. There is simply no way around it.
The cars passing over any given mile upon it must pay for that mile.
Wrong. Highway costs are incurred for the segment as a whole and simply averaged out by mile. Some miles technically cost more than others to build, but accounting for that in a formula that considers total usage for that segment is virtually impossible. So instead we average. If a 20 mile segment costs 100 pesos we divide 100 by 20 and get 5 pesos per mile ON AVERAGE. We then assess the tax incidence based upon (a) the number of cars and (b) the length of their trip. Part (b) MUST be included because trips vary in length - a fact for which there is no way around.
Saying a 30 mile freeway makes more money than a 5 mile one is besides the point.
Nobody's saying a 30 mile freeway makes more than the 5 mile one, less than the 5 mile one, or anything else about it. What I am saying though is that tax revenues for a specific fixed segment of freeway, be it 5 or 500 miles in length so long as that length is identified and held constant in the model, depend upon not only the number of vehicles but the length of their trips. There's no way to escape that fact so it must be included for a model to be valid.
Is the construction being financed with revenue bonds (normal method of construction of public works)? If so, you need to add the interest charges.
The $800 million is the raw number for construction itself. I don't recall the exact financing, though a substantial portion of it is federal returns. But go ahead - add whatever additional costs you desire into the mix. The freeway STILL pays for itself from the gasoline tax well within its lifespan.
Presumably the property underneath the road was once privately owned.
Not since about 1855, if even then. I-10 was built on the old Houston-Beaumont railroad line's Columbus spur. The highway was put in during the 1960's on adjacent right-of-way from the railroad plus some older roadways. As I detailed for you previously, the overwhelming majority of it's expansion is being built upon remaining county-owned right of way, the county-owned stretch of what was left of the RR line, and a county-owned parallel roadway. The only parts they are eminent domaining are a few commercial properties that extend to the corners on existing crossroads and they compose a minimal portion of the loss.
You need to add on the foregone property tax revenue on that land.
No need, because the county does not pay taxes to itself on its own land. What little of your precious tax revenue that may be lost from the tiny fraction of eminent domain acquisitions will be more than offset by the boost to other property value along the freeway when complete.
You need to add in the cost of policing the road and responding to accidents and incidents.
As previously noted, most policing costs on the Katy Freeway are from congestion-induced accidents. The expansion will therefore likely REDUCE policing costs.
Lastly, is the construction estimate in current year dollars, or year of expenditure?
I'll have to look it up to be certain. Either way, the revenues generated are more than sufficient to pay it off well within its lifetime.
I will grant the basic point - freeways in Houston are very heavily used but cheap to construct and therefore generate an atypical amount of money versus what occurs elsewhere in the country.
They are not at all atypical for Texas. I would venture to guess that the same is probably true in most southern and western states save the left coast. The exhorbitant costs you state, if true at all, seem to apply only in the northeast and on the left coast. When considered from terms of geographic land size, that would also seem to make your costs atypical for the nation, not mine.