Civic Data

Data-oriented thinking about where and how people live.

The cost of a road

Leave a comment

[Ed note: It has come to my attention that the original road was actually asphalt, not concrete. I’ll re-run the numbers for an asphalt maintenance schedule in a future post. The short version is that an asphalt road is somewhat less egregious here.]

[Ed. note the second: I’m going to go ahead and just retract the thesis that this road constitutes an example of underfunded maintenance. The maintenance costs for asphalt for this road are probably appropriately annualized at more like $4,000 to $6,000 a year. Compared to a property tax base of $70-100k, that strikes me as pretty reasonable. This doesn’t mean that I’m happy about the suburban pattern of development more generally … but the municipality here was correct to take the free road in exchange for the property tax revenues. If you read onward, carry a big salt shaker.]

Many American municipalities are facing budget crises. 22 have declared bankruptcy in the last two years. Underfunded pension obligations have been the most common cause. I suspect that the costs of maintaining our overbuilt infrastructure — largely unprepared for — will be an similarly important cause of bankrupt cities in the long term. (It’s worth noting that the largest municipal bankruptcy in history — Jefferson County, AL, home to Birmingham — was due to debt refinancing shenanigans surrounding a wastewater treatment plant.)

I wanted to investigate whether the maintenance obligations made by cities make financial sense, after the model of an eye-opening post on the Strong Towns blog. I’ve chosen to focus on a road near and dear to my heart. I’ll call it Waterfront Road — I’ve removed actual identifying information but it might not be too hard to find if you’re curious.

waterfront road

Waterfront Road. The area is gorgeous!

I don’t know the entire history of the road but it was in place as early as 1992 and probably a bit earlier — the earliest deed dates that I can find  in Waterfront tax records are from July 1986. It’s safe to guess that the road was built by the original landowner to provide access to the waterfront. It is now publicly maintained. Conveniently for calculations, it’s almost exactly a mile long, with 20′ of paved roadway and 2′ of drainage channel on each side.

Waterfront Road provides a nice natural experiment because it doesn’t connect through to anything — it doesn’t have any “network value”. In principle, the eastern end could connect through to another nearby road, but due to the circuitousness of the route it still wouldn’t make any sense as a cut-through to anything. The only value of the road is to provide accessibility to waterfront property.

How much does such a road cost?

Construction Costs

The road was originally concrete, and it had one layer of asphalt overlaid in maintenance several years ago. There are many variables that can affect the cost of a road — especially the surfacing materials, but also the costs to clear the right-of-way, width, grade, and soil quality. I tred to find a reasonable range of estimates, but this will be ballpark at best.

Handily, the state of Florida provides a “Generic Cost per Mile Models” pdf for estimating the cost of road construction. The closest estimate they provide is for a “New Construction Undivided 2 Lane Rural Road with 5′ Paved Shoulders,” at about $2,400,000 per mile. This is almost certainly an overestimate — they are presumably describing a road with two 12′ lanes in addition to 10′ of concrete on the shoulders. Accounting for the smaller width of our road (and assuming costs are strictly proportion to the area of road surface) lowers the estimate for our road to $1,700,000 per mile. Furthermore, they are probably describing higher quality construction which can withstand high levels of daily traffic. I’ll treat this as an upper bound on the price of Waterfront Road.

Perhaps a more realistic estimate comes from a fun article on The article describes the plight of property owners who are being forced to pay their share of paving costs for what had formerly been a dirt road. Apparently, the signatures of 9 of 16 property owners sufficed for the county to pave the road and send them all a bill. This road is probably more comparable than the generic Florida road above — it’s a cul-de-sac with no connections except driveways, and the right-of-way is already cleared.

Specifically, each of the 16 property owners is being asked to pay $24,000, for a total of $384,000. This is for initial construction only — the road remains a private road, with maintenance the responsibility of the property owners. Some research — by which I mean playing around with a Google Maps distance calculator — indicates that the road in question is about 3200 ft long and 16 ft wide. Again assuming costs are proportional to surface area, we can scale this up to a mile in length and 24′ in width and estimate a cost of $950,000 per mile.

A final estimate comes from Larimer County, CO, home to Ft. Collins. On page 27 of this document, they estimate that paving a new two-lane road costs $1,850,000 per mile. I’ll assume again they mean something more like 34′ in width, and so we can discount this to $1,300,000 per mile for Waterfront Road.

I’ll be generous and assume that Waterfront, if built today, could be constructed at the low end of the range, or around $1,000,000.

Maintenance Costs

Maintenance is even more complicated. For simplicity, we’ll generously assume that the road will last 60 years, and will need to be resurfaced twice during that time, at years 21 and 41. Going back to the document above from Florida, they calculate that milling and resurfacing the same rural road described above would cost $430,000 per mile. Other estimates are as low as $100,000 — let’s take that for our estimate. At a discount rate of 3%, that’s a net present value of around $55,000 for the initial resurfacing, and $31,000 for the later one. Given our other uncertainties, this is small enough to ignore. I’ll also ignore potholes, every-three-year sealing, etc. They’re real costs but the initial construction is the biggie here.

Total Costs

Thus, our total construction cost works out to something like $1,000,000. If we bond this over 60 years at a municipal rate of 3%, assuming that the road will have to be completely rebuilt in year 61, payments on the road would cost $36,000/year.

Property Taxes

Waterfront Road is the only way to access the lots on the road, and the only people travelling on Waterfront are accessing those lots. This allows us to ask a simple question: how much do the residential lots on the road pay in annual property taxes? The county in question makes the last seven years of property tax records available online, so I looked up the total property taxes paid by the 87 adjacent lots.


Property taxes (in $) collected from all lots adjacent to Waterfront Road.

You can clearly see the local reflection of the real estate boom and bust. Appraised valuations and therefore property taxes peaked in 2007 and have only dropped since then. The peak of the boom in the figure here — 2007 — is a couple of years delayed from the actual 2005 peak of the real estate market, because of the lag time for county reappraisals. I suspect it hasn’t fallen far enough yet: very few lots in the neighborhood have sold recently, only 3 in the last three years. (Based on the tax data: one houses, one unbuilt lot, and a bank repossession of a waterfront lot.)

Now, does this make sense? Residents are paying basically $70-100k in property taxes — and $36k is going towards the road in front of their house.  For comparison, 28% of county revenue currently comes from the property tax, while only 12% of the budget is spent on roads and highways. On the face of things, it sounds absurd.

What’s missing is that the city hasn’t yet had to pay this cost. The initial cost of the road was built into the cost of the property, which was inaccessible until the road was built by the developer. The municipality has had to resurface the road once with asphalt (which itself has bonded annual cost of perhaps $15-20k) but has not yet had to deal with a complete rebuild.

I draw a couple of conclusions from the above:

  1. The road is too wide. This is true independent of whether the city can afford it. This road sees a maximum of 200 cars a day, probably more like 75 to 100. It could easily have been constructed as a wide 1-lane road, with just enough space for two cars to squeeze by each other at a slow speed — 16′ wide or so. (There are currently 40 homes on the block. Many of them — at least a third — are vacation homes and empty much of the time. Many of the others are owned by retirees who don’t drive much. And almost nobody parks on the street.) This would have saved close to 1/3rd of the original construction cost: $300,000. Furthermore, the current design encourages cars to travel above 35 mph — a completely inappropriate speed in a residential zone. This page is a good summary of my thoughts on standard lane widths: essentially, we’ve needlessly (and wastefully, and destructively) overbuilt our roads.
  2. Property tax rates will have to go up sooner or later. The municipality accepted this free road from the developer, and promised to maintain it into the future. In return, it received a large stream of property tax revenue — but also a large, probably unbudgeted-for obligation to rebuild the road 60 years down the line. We barely get to keep half of the property tax money to spend on schools, public safety, sanitation, other roads, and all the other things we expect of our local governments.

Conclusion #2 is not ironclad. The alternative to higher property tax rates is very fast population growth. Population in the county has almost tripled in the last 30 years, and population in the nearby towns has double in the past 10 years. I suspect we’re in the middle of the suburban growth Ponzi scheme, where plentiful new development (like this neighborhood) is providing an apparently free revenue stream, easily paying for the maintenance of the much smaller number of older roads. But if the population ever stops doubling every 20-30 years, there will be hell to pay. (Current population growth for the U.S. as a whole is 0.97%/year, which implies a population doubling time of around 70 years.)

This is not economically sustainable development.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s