Fiscally Sustainable ROI of Highways at Top/Bottom of Market

Thankfully, the thought processes surrounding transportation are getting increasingly complex.  An emerging field of study, analysis, and scrutiny is less strictly about transportation in a vacuum.  We’re no longer thinking strictly how to continually fund and build roads but instead, what is the relationship of the cost of building roads and maintaining them in relation to the land uses around them.  Not only is this critical from a tax base to tax burden ratio, but also from a transportation standpoint.  A better land use plan that requires less trips on the highway puts less burden on that highway and less of a need to continually expand it until there are no destinations left worth going to.

Along these lines, I decided to compare some land uses and their associated tax base with some hypothetical highway costs (either construction or re-construction – it doesn’t really matter) and the time it would take the tax base to pay off this hypothetical tax burden, or highway cost.  To compare, I look at a square mile of land uses along a linear mile of highway and then compare the tax base to various highway costs per mile based on several reported recent project costs.

To add a bit more depth, I looked at two areas, one to the north and one in the south.  Since these areas aren’t fully built out, I projected the assessed values and property taxes over larger areas.  Since a square mile is 640 acres, I cut this in half as ‘net developable.’  The rest would be dedicated to public right-of-way and other necessary tax exempt land uses such as schools and churches that comprise elements of a community.

I used four land uses:  big box, retail pad, single family, and multi-family.  Found local comparables for each land use and then projected those over a quarter of the land area.  For simplicity of the math, I assumed that each of these land uses would be a quarter of the total net developable area.  In other words 80 acres of each.

I then added up the property taxes each of these produced per year to see how long it would take to pay off the cost of a highway project assuming it had a 30 year life-span, in essence, to find what kind of value we must build to determine a positive lifetime ROI and a sustainable development pattern.  I grabbed a range of recent highway costs to compare how long it would take to pay off $90 million/mile, 165m/mile, 240m/mile, and 416m/mile.  Keep in mind, those are all the per mile cost of recent DFW area highway construction costs.

First, we look to the north and the hot development market of Frisco.  Let’s call this the ‘top of the market.’

You can see there are examples of single family, multi-family, big box, and retail pad sites.

highway top

Below I’m going to list land use along with property taxes per acre based on the local comps and then projected over 80 acres.  I also subtracted the hospital and school taxes and only used property tax rates that would go to general fund and things like infrastructure.  Keep in mind, I’m putting the entirety to infrastructure, which we know isn’t really the case.

LAND USE         PROP TAX/Ac./Yr        Over 80 Ac.

Single Fam               $13,325/ac                $1,066,000

Multi-Fam                $9,937/ac                  $ 794,960

Retail Pads               $10,767/ac                $ 861,360

Big Box Retail          $ 5,709/ac                $ 456,720


SUM                                                           $3,179,040

Into 90:  28.3 years – The only solvent

Into 165:  51.9 years

Into 240: 75.5 years

Into 416:  130.9 years – In other words, it would take two human lifetimes to pay off the cost of 30 years of infrastructure.  This is where insolvency comes from.


It’s interesting to see big box retail pays the least in property taxes per acre by sum margin.  Also, I’m burying the lede here but it takes the cheapest possible highway construction cost to have a net positive return on investment within this framework.  The argument against looking at it this way is somebody could suggest that the highway is a public good and it can and should be provided from state and federal tax dollars.  The problem with this narrative is that with diminishing available funds to pay for the second generation of this infrastructure, we need to figure out a politically practical way to either raise funds or design our infrastructure in a more fiscally sound manner.

The other place this argument falls down is asking the question, what is the purpose of transportation and are highway expenditures to this scale (where we are literally taking books out of school children’s hands and lunch off their plates to build more highways) when cities around the globe have proven more efficient ways to design infrastructure, arrange land uses, and move people about without any highways at all let alone less than Texas which has the most highway lane miles in the country.  That’s an awful lot of public expense for a supposedly fiscally conservative state.

Now let’s look at the other side of town:

highway bottom

The image shown above is 635 through southwest Dallas and Balch Springs.  It also has comps for Big Box retail, pad retail, single family, and multi-family.  Again, I deducted hospital and school taxes from the total.

LAND USE         PROP TAX/Ac./Yr        Over 80 Ac.

Single Fam               $ $2,972/ac                $237,760

Multi-Fam                $ 2,390/ac                  $ 191,200

Retail Pads               $ 8,555/ac                $ 684,400

Big Box Retail          $ 7,226/ac                $ 578,080


SUM                                                           $1,691,440

Into 90:  53.2 years – It seems highways don’t pay for themselves.

Into 165:  97.5 years

Into 240: 141.9 years

Into 416:  245.9 years –  Goodness gracious.  Take the most expensive highway we have built in the metro.  Put it into South Dallas, say like Southern Gateway, and it might take over two hundred years to pay off its thirty years of utility.  That’s utterly insane.

What have we learned in this quick exercise?  Only brand new development at top of market is solvent if and only if the highway is built using the very cheapest possible cost.  However, the key word there is brand new.  It must retain value throughout the 30-year span, which conventional suburban development has a very poor track record of actually doing.  It is placeless and generic therefore open to cannibalization by the next newest thing.  Thus, disinvestment and devaluation occurs.  Even in the rosiest of scenarios, this is in all likelihood a net loser.  And we wonder why public sectors go broke building for Growth! rather than actually building places with lasting value that people care about.

As we get closer to the city and the cost of the city increases, we have to be smarter about how to leverage the most amount of tax base and productivity out of highly valuable land.  The land simply costs too much to be devoted to long distanc travel.  Instead, we need to rethink infrastructure to maximize productivity of land use, maximize mobility, while minimizing lifetime infrastructural costs.  I think other cities realized this long ago not allowing highways into their city and investing in transit which is shown to encourage more efficient land use patterns and land value premiums.