Estimating Antebellum Passenger Costs: A Hub-and-Spoke Approach

Abstract

Over 30,000 miles of railroad and 4,000 miles of canals were constructed in the United States between 1815 and 1861. However, the lack of data has prevented the study of this “transportation revolution” for most cities. This article thus enables a closer study of antebellum travel improvements by constructing a hub-and-spoke network that is capable of estimating the cost of passenger travel from New York City and Philadelphia to any U.S. city in 1836, 1850, 1856, 1859, and 1867. The semi-parametric approach provides an accurate cost estimate by using available historical information to determine the travel cost to all other cities.

Keywords:

• passenger costs
• transportation network
• U.S. economic history

Notes

1. The term “transportation revolution” was first popularized by Taylor (1951).

2. In fact, time was not actually standardized across cities before the 1880s, and the official time was often set by each town's clock. In this way, the times provided across cities might be a few minutes off from each other.

3. As time was not standardized across locations during the period, it is unclear how accurate these times were across the cities or to which city's time they referenced.

4. Despite their different publishers, a comparison of Disturnell's 1862 and Appleton's 1859 yields few differences.

5. When stagecoach or wagon routes are listed, the guides often do not contain the travel cost, time, or distance.

6. Because Gorton (1989) examines slightly different years than this study, I have adjusted his costs during the 1850s using a linear function. The network can also be adjusted to incorporate other cost estimates.

7. I account for the additional travel costs of the longer trip, but do not build in shoe leather costs for changing trains.

8. The increase in hubs substantially reduces the distance between hubs and spokes, which in turn reduces the estimation error over time.

9. In general, the choice of the fastest route rules out ocean travel, as even early land routes were faster.

10. See Sinnott (1984) for more details.

11. The lack of comprehensive travel cost data has led redemption network studies to focus on a small number of known travel costs (Weber 2005a), aggregate by state (Gorton 1996, 1999), or focus only on the redemption structure (Redenius 2007). In this way, the example offers an illustration of the network, and enables a comprehensive examination of antebellum note redemption.

12. Inflation calculated from Officer (2008).

13. I only depict the travel costs from New York City because the trends are similar to those from Philadelphia.