Who Will Build the Roads?



Ever since I began my degree in economics, I’ve felt the question, “Who will build the roads?” is equally equivalent in importance as the question of “What is the answer to the ultimate question of life, the universe, and everything?” Instead of the answer being 42, the answer most commonly given is “government.”

If we want to look at it through a theoretical sense, mainstream economics (or at least what I’ve seen in my textbooks) has justified the need for governments to levy tax to build roads, because the private sector apparently tends to not produce the optimum amount of public goods, so government must place an order for infrastructure.

I’m here to burst this nonsense. Essentially, when combining the correct institutional environment for investors and entrepreneurs, the private sector has the ability to provide public goods at an inexpensive and higher quality than government. In particular, I’m going to examine the turnpikes of the 19th century in the United States. For this we will examine the works of Daniel B. Klein and John Majewski. I will introduce other relevant sources that show even in the 21st century we still use private roads, if even in a limited manner. Furthermore, I will provide recommendations on how we can transit towards higher quality roads with the highest cost-efficient results via a unique way of privatizing roads and highways.

Before the Turnpike Revolution

Before the 1790s, most Americans had no experience with any private turnpikes. Turnpikes were essentially built and financed by town governments; typically, those government agencies would enforce a road labor tax. A rather extreme example of this is New York state. New York essentially made eligible males work for a minimum of three days, or they’d have to pay a penalty or fine of $1 ($25.41 equivalent as of 2016). However, as an alternative, these individuals could pay a daily fee to avoid having to build these roads.

This essentially created a weak incentive for anyone to work on public works. Without the profit motive, investments were low and organization of essential resources could be described as slow and unorganized, and we must also take into consideration that in the 1790s the United States was not full-force affected by the Industrial Revolution; the country was essentially still a farming community. This in turn, essentially made crop schedules dictate when roads were built and repaired.

Because organizing labor was completely chaotic in every way possible, road conditions remained inadequate. This was extremely problematic for Americans looking for better avenues to markets. Due to the lack of technology for advanced transportation via river and rail, another solution was needed in order to connect American markets together.

Before the Turnpike Revolution, from 1786 to 1798, some private turnpikes and toll bridge companies did operate and boasted an average annual return of 10.5% in their first six years. Britain had success with private toll roads from 1663 to 1772; the turnpike movement there was so large it received the nickname “turnpike mania.” However, what made the American turnpike revolution more successful was the fundamental ways which these turnpikes operated. British turnpikes operated as non-profits with bonds as the main source of funds, while American turnpikes essentially used stock shares.

The Turnpike Revolution

From the 1800s to the 1830s, nearly 27% of businesses in the United States were turnpikes! This number is around 946 turnpikes incorporated, and by 1845 there were around 1562 incorporated turnpikes, of which 50 to 70 percent had successfully built and operated toll roads. To further examine this success rate, we must look at various state regulations.

New York had turnpike inspections, which were around every 10 miles. This so-called regulation contributed towards 35 to 40 percent of turnpike companies reaching operational statuses. Connecticut essentially sold or gave ownership of its existing roadbeds to turnpike corporations. This privatization resulted in extremely low construction costs, and around 87% of companies reached operational status.

What about government subsidization of such turnpike companies? The only states that subsidized turnpikes were Pennsylvania, Virginia, and Ohio; most turnpikes were financed solely by private stock subscriptions, thus this made subsidization a relatively small factor.

Now we must look at why many individuals invested into turnpikes. Turnpikes then cost around $1,000 ($30,251.20 inflation adjusted) to $2,000 ($60,502.40 inflation adjusted) for around 15 to 40 miles! For the 19th century, this could be considered a costly project. The reason why many individuals decided to hold stock in turnpikes was because of the indirect benefits of ownership. Turnpike stocks were known to make little to no return; however, merchants, farmers, and land owners knew that investing in turnpikes would vastly improve their income, or increase the value of their assets which surrounded new roads. In Pennsylvania, more than 24,000 individuals held turnpike or toll bridge stocks. On average, the stocks were worth $250 ($5,123.19 inflation adjusted). Another reason why individuals and communities wanted to purchase turnpike stocks is because it was considered a civic duty; many towns had meetings to raise funds for new toll roads and bridges.

Below is a table showing the amount of private toll roads built from 1792 to 1845:

Toll Road Incorporations %Successful in building roads Roads built and operated Average Road Length Total Road built. Total cost of Roads (inflation adjusted)
Turnpikes (mostly east coast)  








15,000 miles  


Below is cumulative turnpike investment by state (1800-1830) as a percentage of 1830 GNP:

State Cumulative Turnpike Investment, 1800-1830 (not inflation adjusted) Cumulative Turnpike Investment as a Percentage of 1830 GDP.
Maine 35,000 0.16%
New Hampshire 575,100 2.11%
Vermont 484,000 3.37%
Massachusetts 4,200,000 3.37%
Rhode Island 140,000 1.54%
Connecticut 1,036,160 4.68%
New Jersey 1,100,000 4.79%
New York 9,000,000 7.06%
Pennsylvania 6,400,000 6.67%
Maryland 1,500,000 3.85%
Total 24,470,260 6.15%

Turnpike companies had the ability to build higher quality roads as well. New York State Gazetteer’s Horatio Spafford wrote, “An excellent school, in every road district, and people now work the highways to much better advantage than formerly.” Turnpike companies developed extremely efficient and extremely organized network systems. This in turn allowed competitors, which in this situation all relied on each other, to communicate with one another on essential market factors needed to keep such roads in constant and pristine quality.

Government-built roads were not only worse in quality and shorter, but were more expensive, and took longer to complete. The federal government spent $13,455 per mile ($263,004 inflation adjusted) to complete the National Road, while the Pittsburgh Pike cost $4,805 per mile ($164,292 inflation adjusted). The Pittsburgh Pike was completed before the National Road, and continued to provide higher quality roads. How did a private turnpike corporation provide better quality of roads? The simple answer is that because private entities would charge for the usage of roads, which allowed for a steady and instant amount of cash flow which was crucial for repairs. The government’s method for financing and planning repairs was rather inefficient. Historians have pointed out that travelers generally preferred Pittsburgh Pike compared to the National Road.

Finally, what about cross comparison of international competing nation roads? We’ve briefly discussed British turnpike roads; were American turnpike roads better? In general, yes, again due to how funds were raised. British roads were financed via what could essentially be described as “community bonds,” while American turnpikes sold shares. Because individuals preferred to invest and have actual ownership of such corporations, this essentially allowed the right amount of capital to flow towards American turnpikes, while British turnpikes (although successful) drew less capital.

Modern Day Roads

As of today, many roads are in poor conditions, despite the stimulus package of 2008. A report by USA Today showed that 11% of bridges are in poor conditions, while around 21.32% of modern US roads are in poor conditions. Despite having the 2008 stimulus package passed, President Trump wants to spend one trillion dollars on new infrastructure projects.

Let’s look at the relationship of how much is being spent at both federal and state levels for every mile of road (using the transportation budget, as highways expenditures under the transportation budget is directly linked towards total spending of transportation, correlation of 0.99, see more in the disclaimer section) in each state and compare it to the amount of roads in each state to determine a correlation if an increase in spending actually leads to higher quality.

The results show that, when controlling for multiple outliers, the less government spends per mile of road, the higher the quality of roads built. This result can be interpreted in two different ways:

  • Government is extremely inefficient; expanding spending budgets on infrastructure projects will create more waste.
  • The data set used contains data which is irrelevant, therefore the results shown have either a sampling or non-sampling error.

Interpretation number one is more likely. If government limits infrastructure spending contractors would be forced to use resources more efficiently, thus leading towards a larger amount of good conditioned roads available towards the general public.

Currently in the US there are 2,200 private highways. However, let’s focus on the 91 Express Lane. The 91 Express Lane in California is one of the world’s first private highways and also has the world’s first fully automated toll gates. The customer base is also continuously growing for the 91 Express Lane and it does not suffer from extreme congestion. The reason for this is because toll fees float according to the amount of demand. When there’s a lot of traffic, toll prices increase in order to reflect such demand. Of course, as expected, as prices rise to reflect demand, fewer travelers use the lane (my calculations for the price elasticity demand arc, is westbound 3.10, and eastbound 1.81), and more people travel on the normal 91 freeway.

Solution Towards Modern Day Roads

Before we get into how to fix our crumbling roads, we must also look at what would happen if all highways and roads were privatized. The first question might be, If all roads were to be privatized, wouldn’t the traffic be the same as if they were public?

The answer is, yes and no. Due to the fact that from both historical and current perspectives privately owned roads boast better quality than public roads, this could potentially lead towards improved traffic times, and potentially even better planning of traffic routes. Historically, turnpike operators coordinated routes and other operational issues together extremely effectively. Thus, the most likely result would happen the same if all roads and highways were to be privatized.

Furthermore, while the aggregate amount of time waiting in traffic would be reduced by a small amount, what would happen within those aggregates is that certain highways and roads would charge higher amounts to customers who would pay for convenience of getting to their destinations quicker, while other private roads/highways would only raise their prices by a small marginal amount in order to maximize revenues. In the aggregate, changes in prices would not necessarily influence changes in the amount of traffic (for example, certain studies in Norway conclude that the price elasticity of toll roads there is 0.62), within the actual market of roads/highways, changes in prices may persuade users to take different routes, or potentially use different means of transportation.

How can we fix the modern-day road, bridge, and highway issue in the United States? The answer is to privatize such goods; the ways in which they should be privatized is essential and important. Italy essentially gave control of its highways to one company and their quality has been reported to be less satisfactory than roads in other European countries. What the government should do is divide ownership of roads and highways into many different companies. Let’s assume that every 100 miles of road will be owned by one company; this would form 39,617 new companies. Each company will have a hypothetical 500,000 shares. Let’s assume that each company wants to raise enough capital for five years of repairs and maintenance, while its revenues from tolls would also be sufficient enough to repair and maintain such roads for the remainder of time. From our correlation graph, we’ll use the highest spending per mile of road, and assume that in five years, these costs can be reduced so that we can make an estimate on how much each share would cost.

Year Total repair and Maintenance cost.
1 $1,043,062
2 $917,300
3 $890,110
4 $870,539
5 $772,633
Total $4,493,643
Per Share $8.99/Share

A Discount rate of 4.5% was used in order to predict future expenditures needed.

Once privatization occurs, what would be the return for the American people? I’ve made a graph showing the cost per mile road (repairs and maintenance) from 2011 to 2021 between government provided roads, and with a private oriented solution.

From the graph, I’ve decided to take the assumption that for private sector, it would be inefficient and thus its costs would be higher initially, and that over-time, efficiency increases and thus the private sector over the 9/10-year period would be able to reduce costs. From the logarithmic equation (which is y=-2,076.6ln(x)+10,567.7), we can show that the private market will reach the highest cost efficiency level at year 39/40.

How would private sector roads suddenly find this new kind of efficiency? Breaking up the roads into essentially 30,000+ different companies owning 100 miles of road each would create a huge demand for contractors. Combined with being a new/revived industry, finding essential labor for planning and maintaining, while also installing toll gates, and other factors would generate high initial costs. However, over time, as the industry becomes more efficient and competitive, real spending costs per mile would lower, assuming that no sudden changes happen within the industry.

Compared with government provided roads, the level of spending will always be constant at around $7,000 per mile of road, compared to the lowest cost private sector could potentially achieve of $3,000. It’s not that those who work in government are incapable of lowering costs; it is due to special interests. Contractors who have connections with politicians will be “fairly” awarded contracts to build roads, hence why there’s a large variation of spending on per mile of road between the various states, as well as with quality of roads.

How reliable are our estimates compared to past data and trends? Unfortunately, for private sector roads, finding such data will be tiresome, and will require a lot of coordination and effort.

However, we find that government spending is constant in the past.

Year Spending (inflation adjusted for 2014 dollars) Miles of road Spending per Mile of Road in dollars (10% of spending).
2005 262.99 4,010,000 6,558
2006 268.82 4,031,000 6,669
2007 277.46 4,047,000 6,856
2008 284.06 4,058,000 7,000
2009 298.05 4,064,709* 7,333
2010 315.97 4,067,077 7,769

2009 Miles of Road had to be estimated as no official government information was provided.

With new information, if we analyze spending from 2005 to 2014, and then estimate total spending in year 2020, will we receive the same result? Real spending per mile of road will amount to $7,477 compared to our initial estimates of $7,832 (4.54%). The official figures between 2005 and 2014 show an average real per mile spending of $7,078.

Calculations and Estimates Disclaimer

Due to the lack of information that I can find on actual real spending per mile, the natural thing one should do is make estimates. The estimates use the total transportation cost, divided by length in miles, multiplied by 10%.

Example, 100,000 (total spending)/ 100 (miles of road in the country) = 1000/mile*10% = 100/mile.

The reason why this calculation was performed is because of a direct correlation on the expenditure of maintenance (0.99, in fact rounding it would indicate 1.00). The expenditure of federal maintenance is around 14% of the total budget. For simplicity, we use 10% rather than 14% due to the various efficiencies between federal agencies, and states. However, even using the 14% interval does not change the fundamental analysis results provided above. Using the 10% interval in fact, allowed us to estimate Wisconsin’s cost of maintenance per mile (from actual journal research) of around $3,400 (adjusted for 2011 inflation), compared to $3,800.
















The following two tabs change content below.

Baland Rabayah

Baland Rabayah is a student of accounting and finance at Bangor University and is also pursuing a Graduate Diploma in Economics at the University of London. Baland holds a Diploma in Accounting and Finance from the Bahrain Institute of Banking and Finance. He plans to pursue to continue his studies by doing a master's in economic history and a PhD in Economics. Baland follows a mixture of Chicagoan and Austrian principles in economics, with his influences being Murray Rothbard, Ludwig von Mises, Friedrich Hayek, George Selgin, Lawrence H. White, Peter G. Klein, Ronald Coase, and Milton Friedman. He is currently part of the Being Libertarian Merchandise Project’s management, and runs his own investments. Baland is also the former CEO of MoreTech Bahrain, a start-up company which attempted to launch Bahrain's and the Middle East's first flagship smartphone. He is a racing enthusiast, and regularly races professionally in Bahrain's SWS races.


  1. There’s only one shortest route from a to b. That;s a natural monopoly or a licence to a perpetual free lunch for whomever own the road on that route.

    If their was an efficient way to collect those monopoly profits to fund State service, then yes in principle the provision of private roads might be a good idea.

Comments are closed.