by Iver P. Cooper

version June 11, 2005 (9:06PM)

1. Transportation Infrastructure

1.1 Transportation Network

Table 1-1: Road Construction Costs

Table 1-2: Rail Construction Costs

Table 1-3: Canal Construction Costs

2. Transport Entities (Animals and Vehicles)

2.1 Purchase Cost

Table 2.1.1:Service Animal Cost/Purchase Price/Rental Price

Table 2.1.2: Land Vehicle Construction Cost/Purchase Price/Rental Price

Table 2.1.3: Watercraft Construction Cost/Purchase Price

2.2 Carrying Capacity

Table 2.2.1: Pack Animal Carrying Capacity

Table 2.2.3: Load vs. Pull

Table 2.2.4. Load Pulled by Draft Animals as % of Body Weight, for Various Modes of Transport

Table 2.2.5: Drawn Vehicle Carrying Capacities

2.3. Speed

Table 2.3.1: Speed (General Data)

Table 2.3.2: Speed (Point to Point)

2.4 Feed and Crew Requirements

Table 2.4.1. Animal Feed Requirements

Table 2.4.2 Vehicle Manpower Requirements

2.5 Operating Costs

Table 2.5.1: Wages of Interest in Calculating Operating Costs for Land or Sea Transport

Table 2.5.2: Operating Costs (Food and Lodging)

Table 2.5.3. Ship’s Crew Victualling Requirements and Cost

Table 2.5.4: Miscellaneous Operating Cost Data

2.6. Miscellaneous Comments on Service Animals

3. Transporting Your Own Goods

4. Shipping Goods

4.1. Analyzing Transportation Costs

4.2 Rate per Ton-Mile Data

Table 4.2.1: Cost to Transport One Long Ton of Freight Overland for One Mile by Animal Power

Table 4.2.2: Cost to Transport One Long Ton of Freight Overland for One Mile By Steam Railroad

Table 4.2.3: Cost to Transport One Long Ton of Freight for One Mile By Water

Table 4.2.4: Land and Water Direct Cost Comparisons

Table 4.2.5: Upstream vs. Downstream

5. Purchasing Power Corrections

5.1. Purchasing Power of Money

5.2 London Indexes

Table 5.2.1. Simplified British CPI

5.3. Handling American Data

5.4. Handling Data Stated in Other European Currencies; Purchasing Power Parity

6. Notes on Certain Conversion Issues

7. References

Miscellaneous Notes:

The transportation infrastructure consists of the actual transportation network (roads, rivers, canals, rail tracks, pipelines), and the service facilities at the various nodes (stations, ports, etc.)

The first set of tables estimate the cost of constructing improved roads (table 1-1), railroads (table 1-2), and canals (table 1-3), based on historical data. My methodology is explained in more detail in a later section, but in essence I take the historical cost and deflate it, using an accepted consumer price index for London, to 1632.

It is important to recognize one of the limitations of this approach: it assumes that the cost of the various inputs (labor, materials, land, capital) in 1632 will have the same ratio to their costs in the source year (e.g., 1800), that the CPI for 1632 London did to that for London in the source year. If, say, iron is relatively more expensive in 1632 than the CPI implies, then railroads will be more expensive to build–unless, of course, there is some countervailing factor, such as labor or land being much less expensive.

YearCost Then1632Comments
1792$7,500/mi1349L/miturnpike, 62 miles long, 37′ wide, 24′ stone, total cost $465,000, Philadelphia to Lancaster. (Rae 16)
E19C$2,280/mi271L/mi @2.38s/$turnpike, leveled, and graveled, Hartford & New Haven (Holbrook, 35)
$14,600/mi1737L/miturnpike, Salem-Boston (Id.)
$11,700/mi1392L/miturnpike, Newburyport-Boston (Id.)
$4,500/mi536L/miturnpike, Mass. average (Id.)
1817$9,745/mi1027L/miCumberland to Uniontown, road (Meyer 16)
1817~$12,660/mi1334L/miCumberland to Wheeling, 130 mi., $1,645,679 road (Meyer 16) $1.7mil (Dilts 283). For details, see Meyer 16.
14797d/yd2 road paving, London (Gregory 92)
1816800-1000L/mi419-524L/miTelford style road, England, 30′ wide (Gregory 219)
1914 -263s/yd2 water bound macadam, life 3 years. 65d/ ton mile traffic. (Gregory 293)
7s/yd2 tarmacadam, life 9 years. 72d/ ton mile traffic. (Id.)
10s/yd2 asphalt macadam, life 15 years. 28d/ ton mile traffic? (Id.)
YearCost Then1632Comments
1827$20,000/mi2351L/miestimated cost (Dilts 41)
E19C$2,750/mi Maunch Chunk Rwy (right of way already owned) (Dilts 51)
1836$22,000/mi2923L/miDilts 212.
 $42,700/mi Baltimore to Cumberland, $7.6mil for 178 mi including bridges, B&O (Dilts 309)
 L39,300/mi London & Birmingham 4.4mil L, 112 miles. 4.5 yrs to construct (Hastings 29)
1830s$42,600/mi Boston to Providence, $2,000,000 for 47 mi. (Meyer 323). Meyer 327 says $2,764,296.
1836$20,000/mi Utica and Schenectady. 77 miles, 21 months to build. (Meyer 363)
1839$18,600/mi Raleigh & Gaston, VA. $1.6mil for 86 mi. (Meyer 465)
1847$30,400/mi NY RRs. Utica & Schenectady, 78 mi, $2,265,000, Mohawk & Hudson 17 mi, $1,473,000; Syracuse and Utica, 53 mi., $1,187,000; Auburn & Syracuse, 26 mi., $679,000; Auburn & Rochester, 78 mi., $1,865,000; Tonawanda, 43.5, $754,000; Attica & Buffalo, 31 mi., $307,000; Buffalo & Niagara Falls, 22 mi., $206,000. Total 287.5 miles for $8,736,000. (Meyer 378)
1842$44,000/mi Western RR (Mass.). First section, 54.5 miles built for ~$2,000,000. (Meyer 320)
1849$28,500/mi Western and Atlantic, Ga. (Meyer 457)
1851$28,000/mi Maine RRs., 252 miles, $7,129,192 (Meyer 337)
1850s$37,600/mi Illinois Central (Meyer 536-7, 547)
Horse Railroads.
1856-59$41,200/mi4862L/mi @2.36s/$Boston. For commuters. (Poor 90).
The costs below are of “roads, etc.” or “road & equipment” for regular railroads.
1837-59$33,700/mi4044L/mi @2.4s/$Maine RRs. Average Cost 6,377,177, Average mileage 189. In 1859, Cum Cost was 18,382,207 for 511 miles. (Poor 12).
1842-59$19,500/mi2389L/mi@2.45s/$NH RRs. Avg. Cost 10,686,656, avg. mi. 334.59. In 1859, Cum Cost was 17,626,653 for 546.87 mi. (Poor 40)
1849-59$42,400/mi5194L/mi@2.45s/$VT RRs. Avg. cost 17,645,032, avg. Mi. 416.09. (Poor 71)
1835-59$44,900/mi5478L/mi @2.44s/$MA RRs. Avg. cost 35,237,241, avg. mi. 784.61. (Poor 93)
1832-59$42,900/mi5277L/mi@2.46s/$NY RRs. Avg. cost 48,766,185, avg. mi. 1,136.42. (Poor 228)
1833-59$44,900/mi5523L/mi@2.46s/$NJ RRs. Avg. cost 11,336,752, avg. mi. 252.62 (Poor 381)
1830-59$56,400/mi6937L/mi@2.46s/$PA RRs. Avg. cost 41,677,709, avg. mi. 739.27. (Poor 423)
1830-59$52,000/mi6396L/mi @2.46s/$MD RRs. Avg. cost 14,772,666, avg. mi. 284.08. (Poor 576)
YearCost Then1632Comments
1630s-1660s4.5mil guilder1095L/mi @2s/G658 km (411 mi) Trekvaarten canal system (Hadfield 49).
1733-748,000L4565L/mi8.25-10 miles long, new channel for Dee (Willan 94)
1825$19,231/mi2221L/mi @2.32s/$$7,000,000 Erie Canal, Buffalo to NY (Dilts 25) Albany to Buffalo is 364 mi. (Meyer 191)
1825$13,000/mi Farmington Canal, 83.5 miles long, constructed 1820-25 at cost of $1.1 million. (Guignino; Hindley, 86)
1827$23,573/mi2793L/mi @2.37s/$Champlain Canal.
1826-8$13,000/mi1534L/mi @2.36s/$Oswego Canal. $496,011 for 38.25 mi. (Poor 371)
1830$21,300/mi2599L/mi @2.44s/$Delaware Division Canal $1,275,715, 60 mi, 40′ wide, 5′ deep. (Poor 545)
1840$101,000/mi11161 L/mi @2.21s/$Susquehanna & Tidewater Canal. $4,547,173 for 45 mi. (Poor 557)
1833-45$37,800/mi East Division Canal, in Pa. $1,737,285 for 46 mi. (Poor 559)
1833-45$30,500/mi West Division Canal, in Pa. $3,173,434 for 104 mi. (Poor 559)
1820s$56,900/mi C&O Canal, Md. $10.5mil for 184.5 mi. (Poor 602-4).

In the 1820s, a turnpike could be built for $5,000 to 10,000 a mile, while a canal would be $25,000 to $80,000 a mile. (Guignino; Hindley, 86)

Goods are transported over the transportation network by a variety of physical entities. These include

–pack animals carrying the goods in packs or panniers

–cargo vehicles mechanically propelled by an internal motive power (automobiles, steamships, airplanes)

–cargo vehicles propelled by an environmental force such as gravity, wind or water currents (sailing ships, barges floating downstream, mine cars rolling down an incline)

–cargo vehicles pushed or pulled by humans or draft animals (wagons, barges hauled by tow horses or rowed by crewmen)

–cargo vehicles pushed or pulled by other vehicles (boxcars pulled by a locomotive, barge flotillas pushed by a steam towboart)

The vehicles and service animals may be compared according to any of a number of criteria:

–purchase cost

–rental cost

–operating cost (fuel for vehicles, food and lodging for the animals and crew)

–maximum carrying capacity (load)

–speed when carrying a particular load

–carrying efficiency (load per dollar)

–carrying efficiency (load per crewman)

Sailing ships have a high carrying efficiency (load per dollar and load per crewman) but also a high absolute price. Thus, characters in the 1632 novels are not likely to buy a ship unless they need to move ten tons or more in a single shipment.

TypeYearCost Then(1)1632 eqv.Comments
HorseEliz1-2L Singman 36
Horse1580up to ~40s Emerson 142
Farm Horse160317-18RT Hamburg price in Reichstaler (RT). (Baen’s Bar source)
Flanders mare160823.5L26.8Lthe preferred steed for coaches. Croft 121.
Horse1638up to 83s Emerson,142.
Horse17565-10L4-8Lavg. Pa. 4L-16L. Shumway 44.
HorseEliz12d/day or 2.5d/mile. Singman 36
Horse17762s/day20d/ dayrental proposed by Franklin (Meyer 77)
Wagon, 4H & Drvr177615s/day12.5s/dayrental proposed by Franklin (Meyer 77). This was also in a Pa. 1757 statute. (Shumway 48).
Horse (tow)17946s/day4.5s/ daySchuylkill & Susquehanna Canal. (Meyer 79)
Horse (tow)1825$0.75 /day1.74s /dayPa. (Meyer 225)
Ox19C$25 (Geocities)
Horse19C$75 (Geocities)
Horse 2 reals New World. (Braudel SEL 342)
Mule 9 pesos New World (Salta)(Id.)
Horse1834$456.4LPittsburgh (Atherton 46)
Ox1854$37.504.1LKansas (Settle 38)
Ox1857$37.31 Kansas. 7,740 oxen valued at $288,750 (Settle 50)
Ox1857$44.205LUtah. 1,906 oxen at $84,245.20 (Settle 122)
Mule1859$112.1413.9L543 mules for $60,890 (Settle 97)
Horse1860$17521Lassumed avg. cost of 500 horses for Pony Express (Settle 113)
Horse1853 -59$38744.6L$58,487 for 151 horses of Second Avenue RR, NY. (Poor 268-9)
Horses and Mules1851 -59$18722.4L$87,967 for 330 horses and 140 mules, and harness, of Sixth Ave. RR, NY. (Poor 270)
Horse1863$1109Lavg. cost to Union Army (QMR 9/28)
Mule1863$1058.6Lavg. cost to Union Army (QMR 9/28)
Horse1943-4$300 Italy, max price paid by QM Corps of 5th Army (QMR 3/46)
Mule1943-4$250 Italy, max price paid by QM Corps of 5th Army (QMR 3/46)

(1) L=British pounds sterling as of “Year”.

TypeYearCost Then(1)1632 eqv.Comments
Pomeranian 2 horse coach157642L delivered price in London, with set of spare wheels. (Croft 112).
cart &oxen160034/9/8 Medway area, for cart, 6 oxen, 2 horses. (Willan, 104).
wagon and team175640-60L avg. in Pa. Area. 25L-85L. Doesn’t distinguish 2 and 4 horse teams. Shumway 44.
wagon1834$6 Pittsburgh, to be driven by one horse. Add $14 for the harness. (Atherton 46).
Conestoga wagon1830s$250 Add 6 horses to carry a 3-5 ton load. Dilts 281-2.
Conestoga Wagon1854$190 Kansas. To be drawn by 6 oxen. (Settle 38, 44)
Conestoga wagon1857$19021.2L$122,500 for 645 wagons. (Settle 50)
locomotive1842$7000854LWestern RR (Mass.) (Meyer 320)
Bicycle1880$150-200 (Rae 30)

(1) L=British pounds sterling as of “Year”.

TypeYearCost Then(1)1632 eqv.Comments
river craft
lighter1659250L220LHarwich, 40 tons (Willan, 104)
boat172824L22LLoch Garry, 6 tons (Willan, 104)4L/T
trow1758300L288LSevern, 40-80 tons, includes mast and sails (Willan, 104) 4-7.5L/T
boat174525L26LAvon, at Hungroad, 35 tons, 0.7L/T, add 20L if vessel equipped with mast and sail. (Willan, 104)
barge190665L21LBirmingham 25-30 tons (Willan 104 n2)2.2-2.6 L/T
boat17925s/T4s/TUS, Pittsburgh (Meyer 78)
family boatE19C~$50 US, 30-40′ long, cost $1-1.25/ft. (Meyer 97)
boat1820$100-15024-36sUS, no size stated but probably ~ 20 tons. (Meyer 88)
steamboat1811$38,0004300LUS, first steamboat on Ohio R., 116′. (Meyer 102)
boat 20T17943.75s /day2.8s/daySchuylkill & Susquehanna Canal. (Meyer 79)
ocean-going craft warning: tonnage of merchant ships is burden, or register tons (RT), of warships is displacement tons (DT)
Armed Merch16013700L ~6L/T5300L“Red Dragon”, 600 tons (Milton 73), for spice trade. “Used” price; it was built in 1595 as a privateer. Crew ~200, 38 guns.
Armed merch16011600L ~7L/T2290L“The Susan”, 240 tons, 24 guns, crew 88. Used price; formerly a Levant trader. (British Library website)
ship18C10L/T generic rate for building East Indiamen at Deptford shiyard. (Gardner, 29)
warship17203521L3700L“Blandford”, 20 guns, 364 DT, crew 120-130 (Saville 103)
warship176043,391L44,118 L“Bellona”, 74 guns, 1615 DT, crew 590-650 (Saville 94)
warship178630,232L26,493L“Bellerophon”, 74 guns (3d rate), 1613 tons displacement, 138 ft. keel. Equiv. 18L/DT. (Pope, 37) Cannons, shot and powder weighed ~200 tons. (Pope, 44-5).
warship1799$75,47310,830L“Essex” (USA), 32 guns (5th rate), 850 DT, crew 319 (Saville 118)
warship180148L/ton 12 guns, 177 tons, contract time 4 months. (Armstrong 56)
warship1804350L “Rocket,” (fireship), 62 DT, 4 guns, 8 crew (Saville 123)
warship18XX14L/ton prize value of 32 gun frigate to Admiralty in Napoleonic Wars. (Pope 233)
ironclad gunboat1862  USS Cairo, $101,808 for 512 DT, 175 crew, steam engine, 13 guns, 122 tons of 2″ armor plate.
Merch (wood)180118L/ton general estimate (Armstrong 56)
Merch (iron)1838$30/ton “Rainbow”, first all iron ship of size (750 tons), $22,500. (Armstrong 74, 88)
Merch (comp)186917L/ton “Cutty Sark”, tea clipper with composite hull (iron frame and wood planking). 16,150L for 963 tons displacement. Carried 615 tons tea or 900 tons cotton. (Armstrong 98)
1st Class Merch188414L/ton new ship average (Armstrong 119)
3rd class Merch188410.75L/ton new ship average (Armstrong 119)
steamboat1840  $913. “Thomas Wilson (Poor 557)

(1) L=British pounds sterling as of “Year”.

Animals can be used as pack animals (which carry the load directly, usually in panniers) or as draft animals (which pull a loaded vehicle).

Table 2.2.1 sets forth the carrying capacities of pack animals.

AnimalCarrying Capacity
human porter50-60# (Landels 171). 100# for men, 80# for women. (Skeena)
pack mules (BW 600-900#, Landels 172)30% BW (25% on hilly ground)(Landels 172) 20-26% BW. (HNC) 33% BW. Avg. mule used by QM Corps in WW II was 1000-1200# BW. Cargo pack saddle weighed 100#, So cargo load was ~200#. (Thrapp) carries 300-350# (Savory) carries 300-350#, for 20-25 mpd. Forced march of 280 mi. in 3 days with 200# load. Another of 108 mi in 16 hrs. Another of 85 mi in 12 hrs. (Savory)
donkeys (BW 300-600#)small donkey can carry 120#, large one like small mule. (HNC; Landels 172)
Pack/light horse (900# BW avg, 600-1200# range)Feed-fed horse on moderate or better road can carry 25-33% BW at walk for 10 hr. day. Grass fed horse can carry 15-20% BW for six hours in wilderness, rest of time needed for grazing. (HNC) 250# (Olsen 175). 350# (Emerson 249). at least 168#: two bushels of alum salt, each 84#. (Meyer 119)
Camels450#. (Samhaber, 96). 0.4-0.5 tons. (Braudel SEL 343).
Goats BW 200-250#50#. (Harwood)

Many animals can be used more effectively as draft animals; they can pull a heavier load than they can carry. The draft capacity basically depends on friction. In general, it is harder to slide an object than to roll it.

The next table describes the pull which different draft animals can generate. Some explanations will be helpful. Technically speaking, the weight of an object is not its mass, but rather the gravitational force which acts upon it. Force is mass times acceleration. Since the gravitational acceleration is virtually constant anywhere on the earth surface, if two objects are equal in mass, they are also equal in weight.

The table expresses the average weight of various draft animals in force units (kilo newtons). One newton is the force needed to give a kilogram mass the acceleration of one meter per second per second. Gravitational acceleration is about ten meters per second per second, so an animal with a weight of one kilonewton has a mass of 100 kilograms (about 220 pounds).

Multiplying the average weight by the pull-weight ratio, we get the average pull, which is the force the animal exerts on the load.

When a draft animal exercises a force causes a load to move, it does work. Work equals force times distance. Power is a measure of how quickly work is done, it is work divided by time. It can also be thought of as force times working speed, as in the table below. Power is measured in watts, which, technically, means newton-meters per second.

The energy output is the power output times the time, and is a measure of the work done by the animal. One kilowatt-hour means the energy produced or consumed by one kilowatt of power output over the course of one hour. One joule is the energy associated with one watt of power over the course of one second.

Table 2.2.2. Draft Animal Performance
AnimalAvg Wgt kNPull -Wgt RatioAvg pull NAvg Wrkg Spd m/sPower Output (W)Wrkg Hrs / DayEnergy Output /Day (MJ)
Ox4.50.11S1: 495 S2: 535S1: 0.9 S2: 0.7S1: 450 S2: 391610
Horse4.00.13S1: 500 S2: 535S1: 1.0 S2: 1.1S1: 500 S2: 5871018
Donkey3.00.13S1: 200 S2: 134S1: 1.0 S2: 1.1S1: 200 S2: 14743
Mule3.00.13S1: 400 S2: 267S1: 1.0 S2: 1.1S1: 400 S2: 29368.5
Human  S2: 80S2: 0.8S2: 61  

Source: (S1) Table 4, Hicks, Alastair, “Power and Food Security,” SD Dimensions (October 1997), paper presented at the Int’l Solary Energy Soc’y 1997 Solar World Congress, Aug. 24-30, 1997. Online at ; citing Tools for Agriculture, 1992. (S2) Watkins, “Comparison of Power Produced by Various Draft Animals and Humans,

I have some more detailed information (Dale Wagner, “Loads”, on Draft Horse Chat Board.) on draft horses:

1200# horse can exert pull of 180# (15% BW) for ten hours, and can briefly pull 1800#.

1600# BW horse can exert pull of 200# (12.5% BW) for ten hours, and can briefly pull 2400# (150% BW).

2000# BW horse can exert pull of 220# (13.75%) for ten hours but can briefly pull 3000#. (good for starting a heavy load).

Note that the “Conestoga horse” had BW of 1600-1700#. (Shumway 137).

Another source (I unfortunately misplaced the note which identified it) said that a hose can pull 100-150#, avg. 120#, for 8-10 hours, at a working speed of 2.5 mph.


When an object is pressed into a surface as a result of the force of gravity, a frictional force is generated which opposes any attempt to slide or roll the object along the surface. That force is proportional to the component of the gravitational force (weight) which is perpendicular to the surface. To slide or roll a loaded vehicle, a draft animal must overcome the frictional force generated by the combined weight of the vehicle and its cargo. Generally speaking, less force is needed to cause a wheeled vehicle to roll than to cause an equally weighted sled to slide. Sliding friction is the result of surface asperities catching and releasing, while rolling friction arises from the deformation of the wheel and the road surface.

The ratio of the frictional force to the force pressing the object against the surface is called the coefficient of friction.

The coefficient of sliding friction is, as expected, much greater than the coefficient of rolling friction. Also, the force needed to start a stationary object sliding is greater than that needed to keep a sliding object moving, and so the “static” coefficient is greater than the “kinetic” one. For example, wood on wood reportedly has a static coefficient of 0.25-0.5, and a kinetic coefficient of 0.2. (Center for Advanced Friction Studies). The most relevant values for sledding are those for waxed wood on wet (0.14 static, 0.1 kinetic) and dry snow (0.04 kinetic)(

If the motion is uphill, the necessary pull must overcome both the frictional force, and the component of the weight which is parallel to the surface, and thus is equal to

W (k cos theta + sin theta),

where W is the weight (in units of force), k is the relevant coefficient of friction, and theta is the angle of inclination.

On the other hand, if the motion is uphill, then the gravitational force is assisting the movement, and the necessary pull is

W (k cos theta – sin theta).

On a horizontal surface, sin theta is zero, and cos theta is one, so the pull needed is simply kW. The maximum load W which can be moved by a given pull is then 1/k.

There is actually surprisingly little published data on coefficients of rolling friction. Elert, “Rolling Resistance,” The Physics Hypertextbook, gives the following values for bicycle tires: on wooden track (0.001), on smooth concrete (0.002), on asphalt road (0.004), on rough but paved road (0.008). Allan at Colorado State says in “Vehicle Power Requirements” that a typical coefficient of rolling resistance for cars is 0.015. Urieli at Ohio University, in an article on scooter mechanics, gives a value of 0.01 for rubber tire on tarmac.

David Lawyer, in “Train vs. Auto Energy; Vehicle Resistance,” says that the rolling resistance is 0.01 for auto tires on roads, 0.007 for truck tires (higher pressure) on roads, and 0.001-0.002 for railroad steel wheels on steel rail. The rolling resistance is less for steel on steel than rubber on tarmac because rubber tires flex more and the deformation absorbs energy.

“Vehicle Motive Power, MTU13.HTM” gives the following values for average rolling resistance of rubber tires: on dirt roads (.028), on crushed gravel (.02), on flexible pavement (.013), on rigid pavement (.009), on ice (.015), on mud or on dry sand (.1), and on damp sand (.075). For snow, figure .04.

Some of the sources on overland transport do give a load-to-pull ratio for a particular wheel on a particular road surface.

SurfacePull XSource
wagon wheel, probably iron-tired, on …
earth road, level, avg30Gregory 125
earth road, level, bad10
earth road, level, good37.5
asphalt, good30-60Gregory 126
wood, good10-15
Iron-tired wheels on …
sand, deep and loose7Lay, Ways of the World, Table 3.1, p. 89, cited in
graveled earth15 
macadam, worn or little used20 
broken stone on earth35 
Solid Rubber wheels on …
“reasonable” surfaces45 
broken stone on paved foundation, sheet asphalt, plank road50 
Pneumatic Tire wheels …
“reasonable” surfaces60 
pavement, macadam70 
best pavement180 
steel plate or stone trackway250 

2 horses working together, subtract 5%

3 horses working together, subtract 15%

4 horses working together, subtract 20%

(Gregory, 125)

The Historical Novelists’ Center (HNC) has published information concerning drafting ability for equines, relative to their body weight. The data on oxen is from Landels 177.

Load onHorseMule or DonkeyOxen
wheeled vehicle; good road; walking pace100%120%150%
car on rail track1000%1200% 
barge on water6500%7800% 

This data has the following implications:

1) a mule or donkey is about 20% stronger than a horse of equal body weight, and that an ox is about 50% stronger. (The data is not consistent with that of table 2.2.1, which gives them equal pull-weight ratios);

2) the coefficient of friction between wheels and a “good road” is about ten times that between railroad car wheels and a rail track;

3) the coefficient of friction between wheels and a “good road” is about 65 times that between a barge and water (the latter is really a drag coefficient); and

4) the coefficient of friction between wheels and a “good road” is about half that between the sled and the surface contemplated.

Implication #4 makes sense only if the sled is on snow or ice.

It is interesting to note that an 80 pound sled dog can drag a 40 pound load on a travois. (sled dog manual) That is 50% of its body weight. Sled dogs are used mostly in arctic regions, so it is likely that the sled is on snow or ice.

wagonup to 6T?6 horses could draw wagon carrying six tons of goods (legal limit)(Olsen 175)
Pittsburgh wagon1.5-2Tintroduced in 1824. Drawn by 6 horses or mules. Preferred wagon train size was 25 wagons.(Settle 8) Count of Paris said that six mule American wagon carried load of 2000# (Huston 216).
Conestoga wagon3-4T(Settle 8, 43). Load of 2.5T more common. (43) In 1857, 48 trains made of 645 wagons, avg. train thus 13.4 wagons. Carried 3,870,797#, so avg. load 6000#. (Settle 50) Huston suggests that the ones used for military supply were drawn by 10-12 oxen. (156).
Murphy, Espenshield, Studebaker wagons5TUsed in 1850s. (Huston 156).
early RR freight cars6T-20T6T: 1863. Union Army (QMR 9/28). 10T or 40 men: late 1864. Sherman’s army. Ran four convoys a day, each convoy four ten-car trains, speed 10 mph, for total capacity of 1600T/day. (Huston 207, 208). Needed 4# for each man (3# rations and 1# ammo) , 26# forage for each horse, 23# forage for each mule, per day. (Huston 214). 20T, 28′ long, two 4-wheel trucks, Civil War era. (Boyd 96).
later RR freight cars 50T: 1920s, 40′ long. (Boyd 98).

My speed data falls into two categories: generic speed information intended to give a general sense of how fast a particular mode of transportation was, and speed data specific to a particular route.

Land Animals and Animal Propelled Vehicles
Foot courierRoman imperial couriers, 20 mpd (Gregory 61) 1635 footpost, 16-18 mpd (Crofts 49) England, “running footmen”, record of 148 miles in 42 hrs. with one overnight stay (Laughan, service of Lord Berkeley) (Crofts, 52-53)
Pack muleCan go 25 mpd on good flat terrain with good weather. 5 mpd in mountainous region. Avg speed 15 mpd. (HNC) 15-20 mpd with good roads/weather (Emerson 249) 50 mpd on open ground; 3+ mph but only needs 4-5 hrs sleep/day. (Landels 172) 50 mpd if lightly loaded, on level ground (over 3 mph)(geocities). Carries 300-350#, for 20-25 mpd. Forced march of 280 mi. in 3 days with 200# load. Another of 108 mi in 16 hrs. Another of 85 mi in 12 hrs. (Savory)
Pack horseat least same speed as pack mule (HNC).
Pack GoatSpeed 15 mpd. (Harwood)
DonkeyAvg. speed 10 mpd. (HNC)
oxen pulling wagonOregon Trail, 15 mpd (8-9 hrs./day travel time, thus ~ 2 mph). (Bill Speiden, to Geocities site). 1 mph (Landels 177).
chariotTiberius, 3 relays of chariots, 200 miles in one day. (Caesar did 100 miles in a day, with night travel.) (Gregory 61)
Horse and rider, without staging20-30 mpd (Emerson 247)
post riders, EnglandEliz. 10 mph, 100-160 miles per day. (Singman 89). 7 mph summer, 5 mph winter with reasonable horses (Crofts 81) 7 mph summer, 2 mph winter (Walker, 65) 1817, 7-10 mph(Gregory 196)
post riders, France1817, 5-6 mph (Gregory 196)
Horse-drawn coachpublic coach (1650s) was drawn by 6 horses. Typical speed in England, outside London, 29 mpd (if day=8 hrs, then 3.5 mph). Flying Coaches (1657) traveled 12-14 hours, but speed was the same. (Croft, 123, 126)
Conestoga wagon1750, 15-20 mpd, 2 mph. 3 ton load, wide wheels, pulled by 4-6 horses, (Rae, 13) 8300# load carried at 18 mpd (Shumway 73) 15-20 mpd (Shumway 157)
large wagon, 4-6 horses1805, Carolinas, 20-24 mpd. (Meyer 82)
wagon, 4 horses1830s? 30 mpd (Meyer 302)
Self-Propelled Land Vehicles
train18 mph B&O Tom Thumb (Gordon 27)
train13 mph Stephenson rocket 1829 drawing 12.75 tons (Hastings 14)
atmospheric (pneumatic) RRup to 70 mph (Hastings 30)
Sailing ship 
Canal and River Boats10 mpd. Thames barges sailed upriver. (Emerson 254-5). 4-5 mph(Gordon 26) 18C. 2.5 mph (Willan 102). German river. 12 km/day (Braudel II, 357)
steamboat, 50-400 tons1817, 60 mpd upstream on major river (Ohio or Miss.) (Meyer 116)
~1500?ship, Alexandria to Venice, 65 days avg (min 17 days, max 89 days)(Kohn I)
?ship, Constantinople to Venice, 37 days avg (min 15, max 81).(Kohn I)
1694 boat, Provence to Lyons, 6 weeks. (Id.)
1700 12/24/1700 Chusan -> 7/1/1701 Portsmouth. (Armstrong 22)
1709 boat, Loire Nantes to Orleans, 3 months. (Id.)
SAIL ship, Acapula to Manila, 3 months, return is 6-8 months (Braudel I, 419)
 6 mphship. Cross Atlantic, Rotterdam to West Indies, one month. (Armstrong 71)
1817 English Channel to Bombay, 82 days, “Thomas Coutts” (Armstrong 58)
1838 NY to Canton 92 days, return 88 days. (Armstrong 89)(???? iron ship “Rainbow”)
1840s clipper “Sea Witch”, New York to Hong Kong, 104 days, Canton to NY 81 days. Maintained 275 mpd for 10 consecutive days. Best day was 289 mpd. (Armstrong 74)
  clipper “Flying Cloud”, 1793 tons, avg. 222 mpd, best run 427.5 mpd. 89 days NY to SF. (Armstrong 91).
1849 clipper “Oriental”, 1003 tons, 81 days NY to Hong Kong, avg. 200 mpd. (Armstrong 93)
1869 clipper “Cutty Sark”, 921 tons, 7678 miles in 30 days (Armstrong 99)
 465 mpdrecord set by clipper “Champion of the Seas” (Armstrong 98)
Inland Waterway
  horse drawn barge, Lyons to Avignon (downstream) 24 hours. (Braudel II 357)
161913-17 mpd3-4 days to tow boat 50 miles on Welland. (Willan 102)
18075 mphsteamer, New York to Albany, 32 hours out, 30 hours back. (Armstrong
1817 steamboat, Falls of Ohio to New Orleans, 1500 miles, 9 days (2 on sandbar)(Meyer 115)
1817?ordinary boat, Louisville to New Orleans, down 28 days, up 90 days. (Meyer 115)
1817?steamboat, Louisville to New Orleans, down 12 days, up 36 days. (Meyer 115)
1817?100 ton barge, New Orleans to Mayville, 1730 miles, three months poling up (14-16 men), return downstream in 20-25 days. (Meyer 115)
E19C?poled, hauled or sailed boat, New Orleans to Pittsburgh, 100 days (Meyer 107)
1815?Savannah to Augusta. Poleboats 14 days, wagons 7 days, stagecoach 3-4 days. (Meyer 255)
1820?steamboat, New Orleans to Pittsburgh, 30 days. (Meyer 107)
1820?steamboat, Louisville to New Orleans and return, 41 days. (Meyer 107)
1832?boat (steam?), Pittsburgh to St. Louis, 8d (Atherton 90)
Land – Post Riders
 7 mph 63 mpdpost rider, Prince Henry, Richmond to Huntingdon, 63 miles in 9 hrs.(Crofts 86-7)
 7 mph 49 mpdpost rider, Master of Ceremonies, Dover to Gravesend, 49 miles in 7 hrs (Id.)
15837.9 mpd 5.5 mpdLondon to Berwick (330 miles), 42 hours in summer and 60 hours in winter. (Walker 81, 99). Summer considered to be Annunciation to Michaelmas.
16017 mph 160 mpdpost rider, Roger Boyle, 320 miles in 46 hrs. (Id.)
1603133 mpdpost rider, James Carey, London to Edinburgh, 400 miles in 3 days (carrying news of death of Queen Elizabeth). (Walker 97)
15868 mph 110 mpdpost rider, Richard Hawkings, 110 miles in 14 hrs. (Id.)
1595-16044.7 mphavg. speed for letters on Exeter to Hartford Bridge, 130 miles. For Dover to Dartford, avg. was 4.6 mph. Gen’l avg. was 4.6 mph. Includes night travel. (Id.)
167832 mpdRoger Keate, Weymouth to London, 128 miles in four days, stopped by weather. (Walker 74)
177675 mpdpost rider, Omsk to Tomsk, 890 km in 178 hrs. (Braudel I, 416)
 5 mphYork to London (200 miles). (Walker, 172)
Eliz40 mpdby horse, 120 miles, Stratford to London (Emerson 247)
16542.2 mphKing’s Lynn to Cambridge (~44 miles) by horse, in 20 hrs. (Willan 102)
168115.6 mpdLouis XIV, by coach, Versailles to Bourbon l’Achambault, 156 miles in 10 days. (Tarr 233)
L17C25 mpdMme. De Sevigne, by coach, to Rocher (Id.)
176323.5 mpd1763, London to Edinburgh, ~400 miles, 17d (Meyer 51)
1810?13 mpdFort Schuyler to Schenectady, 118 miles, 9 days, land. (Meyer 174)
181265 mpd1812, Paris to Geneva, 390 miles, 6d (Meyer 51)
181261.5 mpd1812, Paris to Basel, 369 miles, 6d (Meyer 51)
181273 mpd1812, Paris to Strassburg, 366 miles, 5d (Meyer 51)
1813?summer road, Boston to Baltimore, 26d . (Meyer 59)
?“, Baltimore to Richmond, 10d (Id.)
?“, Baltimore to Augusta, 33d (Id.)
?“, New York to Augusta, 50d (Id.)
?“, Philadelphia to Augusta, 45d (Id.)
181656 mpd1816, Philadelphia or Baltimore to Pittsburgh, 310 miles, 5.5d, by stage. (Meyer 51)
1816?Philadelphia to Quebec, 7 days (Meyer 51)
181712-20 mpdConestoga wagon, Philadelphia to Pittsburgh, 16 days (Atherton 82)
181734.5 mpdsummer road. Coach. Richmond to Fredericksburg 69 mi, 2d, (Meyer 60).
181911.5 mphCoach, NY to Philadelphia, 8:35. 11.5 mph. (Meyer 63).
182916 mpdoxen-drawn wagon. Santa Fe Trail. Round Grove to Council Grove, 102 miles in 6.5 days. Round Grove to Caches, 357 miles in 22 days. (Settle, 11-12).
18346-20 mpd1 horse wagon. Pittsburgh to Illinois. (Atherton 46)
184123 mphBoston to Norwich, 104 miles, 4.5 hrs. (Meyer 321)
1840?24 mphBoston to Providence, 47 miles, 2 hrs. (Meyer 323)
1840s?Stage van, London to Birmingham, 30h (Hastings 35)
1840s?Stage wagon, London to Birmingham, 60h (Hastings 35)
1840s?Canal barge, London to Birmingham, 3h (Hastings 35)
1840s?Rail, London to Birmingham, 5.5h (Hastings 35)
18513.7 mphroad Vicksburg to Jackson, 46 miles, 12.5 hrs in summer. (Meyer 475)
185118.4 mphRR Vicksburg to Jackson, 46 miles, 2.5 hrs. (Meyer 475)
E19C3.7 mpdwagon on ordinary road, Baltimore to Wheeling, 265 miles, 5wk (Dilts 283)
E19C9.3 mpdwagon on national road, Baltimore to Wheeling, 265 miles, 2 wk (Id.)
183118 mpdwagon, Wheeling WV to Baltimore, 266 mi. , 15 days (Shumway, 73, cites “Jonathan Knight”)
1830s13 mpdwagon (Daniel Barcus), Baltimore to Mt. Vernon, OH, 397 mi., 30 days (Shumway 73)

These are major components of the operating costs, which are discussed in the next section.

Oxjust grazes.
MuleUS Army fed mules 9# oats and 14# hay per day. (HNC) Heavy draft mule fed 12# oats, 16# hay and 8# straw/day. (Savory) 1863 daily ration for horse or mule is 12# hay and 10# grain. (QMR 9/28)
HorseUS Army fed 900# horse 12# oats and 14# hay per day. HNC says general rule is 1.25-1.33# grain and 1# hay per 100# body weight for horse doing heavy labor.(HNC)
YearCrewTons /CrewVehicle
1669812.5100 tons Dutch(Braudel III, 190)
16964-55-620-30 tons, French ship (Braudel III, 191)
2-310“, Dutch, 10T/m (Id.)
10-12 150-200 tons, French Ship (Id.)
7-8 “, Dutch (Id.)
18 250, French (Id.)
12 250, Dutch (Id.)
20 300, French  (Id.)
16 300, Dutch  (Id.)
2516400, French (Id.)
1822400, Dutch (Id.)
19C  pack mule or pack horse train, one packer riding horseback for every ten pack animals. (Skeena)
1855312.4wagon train of 25 wagons, crew is 30 teamsters plus wagon master. Tons/crew assumes 3 ton loads. (Settle 44)

All modes of transportation require some form of human crew, and these must be paid a wage.

JobYearWage Then1632 eqv.Comments
Land Travel
footpostlate 1500s~20d/day 13-15s for run of 240 miles, Bristol to London and back. (Gregory 154). Speed was probably 15-18 mpd, so effective pay was ~1s/day. Walker 118 mentions a 6d/week retainer.
Post rider1557+20d/day English “posts in pay”, all roads except Dover-London. Required to keep 3-4 horses. Eventually, became postmasters rather than riders. (Crofts 65, 70). See also Walker 91. Earlier, Walker gives rate of 1-2s/day (45)
laborer15941d/day summer (double in winter), plus meat and drink (Emerson, 143).
Road laborer174312-16d/day London area. (Gregory 160, 162)
carter18C8-9s/wk Oxfordshire, late 18C (Olsen, 140)
agric laborer18C6-8s/wk Olsen, 140
wagon driver17562.5s/day Shumway 45
wagoner177920L/day! Valley Forge inflationary price. (Huston 63)
wagoner1825$0.75/day Pa. (Meyer 225)
Wagon teamster1855$25-30/mo plus board. (Settle 44)
Wagon master1855$125/mo plus board. (Settle 44)
Coach driver1859$40-75/mo plus board. For Concord public 6 passenger coach (Settle 97).
Teamster1863$25-30/mo (QMR 9/28)
Mercenary Wages (possible caravan guards)
Merc Pvt.16304.25RT/mo. RT=Reichstaler. Baen’s Bar source.
Merc Cpl.16308RT/mo. RT=Reichstaler. Baen’s Bar source.
Merc Sgt.163014RT/ mo. RT=Reichstaler. Baen’s Bar source.
Sea Travel
French sailor169612-20 livres/mo Braudel III, 191
Dutch sailor169610-12 livres/mo Braudel III, 191
1st mate18C5L/mo Mid-century, East Indiaman. Olsen, 143
sailor 24s/mo Olsen, 143
common seaman17001.32L /mo Merchant Shipping Industry. median 1.23. (Rediker 304)
seaman1700 -501.66 Merchant Shipping Industry. Mean (Rediker 306)
Qtr. master2.06 
2d mate2.97 
crew of whaler1772  monthly pay. Mate, carpenter and surgeon, each 3L10s, 2d mate and carpenter, each 2L10s, seamen each 1L10s, + bonuses. (Armstrong, 38).
River and Canal Travel
boys pulling boat17C6d/day Willan 106
waterman174518s/wk Bristol Avon. Willan 111.
Bargeman17945s/day Schuylkill & Susquehanna Canal. (Meyer 79)
Bargeboy17944s/day Schuylkill & Susquehanna Canal. (Meyer 79)
Bargeman1825$0.75/day Pa. (Meyer 225)
Bargeboy1825$0.50/day Pa. (Meyer 225)
laborer (canal bldg)1825$0.30/day Meyer 287
laborer (canal bldg)1827$12-15/mo Meyer 288

In the seventeenth century, the employer was also usually responsible for feeding laborers; the workers did not pay for the food out of their own salaries. For travel on the employer’s business, the employer must pay for lodgings, too. Naturally, service animals also require feed and stabling.

ItemYearCost Then1632 eqvComments
1 meal at innEliz4-6d/day Singman 36
food to goEliz4d/day Singman 36
bed innEliz1d/day Singman 36
lodge horseEliz12-18d/day Singman 36
bed in ginshop17512d/day Olsen Table 12.4
bushel of oats175620d Pa. (Shumway 45)
hay for one horse17563d/day Pa. 10s for 10 nights hay, 4 horses. (Shumway 45)
hay, 1 cwt17573s6d Pa. 1757 statute. (Shumway 49)
Hay for 1 horse8d 
bushel of oats3s 
bushel of Indian corn or rye3s6d 
3 meals~18004.5-5.5s West of Alleghenies. 2-2/6 each for lunch and dinner, 18d for breakfast. Meyer 74
3 meals and lodging~1800$2.25 NE USA. 50 cents for breakfast and lunch, $1.00 for dinner, 25 cents for a bed. Meyer 74.
horse and rider1817$1/day breakfast and dinner for rider, feed for horse, lodging for both, in Virginia. Meyer 75.
Meal and bed18172s4d/day for coach passengers. Meyer 75
3 meals and lodging18174s6d NE USA. 2s3d for meals and 2s3d for lodging. Meyer 75.
Food & lodging for rider and horse1817?$1 37.5 cents for breakfast and horse feed, 12.5 cents for feed at noon, 50 cents for supper and lodging for both. Meyer 75.
food & lodging for wagoner & horsesE19Cunder $1.75 At wagon stand. “X8 for US1960 equivalent”. (Shumway 92) 6 horse team if Conestoga.
Hay1863$25/ton Union Army (QMR 9/28)
Corn1863$1.25/bshl Union Army (QMR 9/28)

Oxen can subsist entirely on grazing along road (Settle 11).

All of the data I have seen on the cost of food for travelers has been for those staying at inns. That leaves open the issue of what the victualling costs for ship crews might be. I have used the official weekly ration for the Royal Navy in the early nineteenth century to estimate food consumption by ship crews, and then used a variety of sources to come up with a total cost for the standard ration. For the ration itself, see Dudley Pope, Life in Nelson’s Navy, 150-9.

7# bread (biscuit)Singman says load of bread cost 1d says one loaf is about 1#.7d
7 gal beer or 3.5 pts rum or 7 pts winePope gives an 1800 price of 8s/gal for rum and 5s/gal for wine. The Talber price for beer is 1d/quart, which implies 4d/gal. Olsen 1 gallon is 14d.28-98d
4# salt beefPope gives 2d/#, or 8d. Talber has 14# for 1s, implying 4# for 3.5d. Higginson has 112# for 18s, implying 4# for 8d.8d
2# salt porkPope gives 3d/#, or 6d. Higginson has 100# for 25s, implying 2# for 6d.6d
2 pts peasAt George Race’s shop at Westgate (1856)’s+Shop peas are 5d/quart, 2 pints=1 quart, so 5d. Based on the butter and cheese data, price in 1600s was probably 1/3 to 1/4 that.1.25-1.67d
3 pts oatmealTalber quarter (28#) of oats is 16s. So 1# is 0.57s or ~7d. Sounds too high. George Race has oats for 1d/#, and meal for 30d/stone or 2.14d/# (1856) one quart corn meal is 1.4#. So 3 pints weighs 2#. So Race price would be 4.3d. But note comments on butter and cheese prices.1.08-1.43d
6 oz butterTalber has 1# for 3.5d, implying about 1.3d. Higginson has 24# for 8s, implying 1# for 4d, and 6 oz for 1.5d. Singman says 1# butter is 3d. George Race (much later) had 1# for 12-13d. Olsen 1# is 10d (1780,90s)1.1-1.5d
12 oz cheeseTalber has 1# for 2.5d, implying 1.8d. Higginson has 56# for 12s, implying 2.57d/#, or 1.9d for 12 oz. Singman says 1# cheese is 1.5d, so 12 oz is 1.2d. George Race much later had 1# for 8.5-9d. Olsen (1787) 6# for 28d, so 1# for 4.67d.1.2-1.9d

Note that weights are not adjusted for the purser’s measure because we are calculating the provisioning costs, as opposed to the actual diet of the sailor. The sailor received 14 oz. pounds of bread and meat, 12 oz. pounds of butter, and 9 oz. pounds of cheese.

The beer served was “small beer”, i.e., not strong. Beer was replaced with wine in Mediterranean, and with rum or wine in the West Indies. The rum was watered down, 3 parts rum to one part water.

All in all, looks like about 60d a week, call it 9d a day per sailor.


I also have operating cost data which does not fit neatly into a particular category.

ItemYearCost Then1632 eqvComment
towed 4 miles by horse17231s Willan 101
hire horse for towing16C  25 sols tournois a day, plus 10 sols for feed. (Willan 101)
cart & oxen160037L/yr Medway area, for cart, 6 oxen, 2 horses. 27L/yr. for pasture hay and incidentals, and 10L/yr. for servants’ wages. (Willan, 104).
riding postElizspec 3d/mile + 4d/stage for guide or 12 first day and 8d/each day after. (Emerson 248)
riding postL16C E17C2.5-6.4d/mi 1584. 2d/mile for each of two horses (rider and guide), + 1 groat (4d)/stage (10 miles) for tip to guide London to Berwick 93s10d London to Chester 47s8d London to Bristol 37s London to Plymouth 61s8d London to Dover (flat rate per stage) 26s8d By end Eliz, 3d/mile In 1635, allowed to dispense with guide and second horse, 2.5d/mile for horse, leave at next post. (Crofts 84)
riding post   Timothy Hutton, Scrooby to Tuxford. 10s for horses and guide, 7s10d for candle supper and breakfast. Scrooby to Doncaster, 7 miles. 9s, plus 2s3p for meal. (Walker 92)
operating steamship1878  Liverpool to Bombay and back by way of Canal of Port Said (Suez). Wages 817L, provisions 565, fuel 1110, stevedores’ and port charges in Liverpool and Bombay 1045, canal dues 1700, insurance 768, advertising and commissions 465, management and office 232, incidentals 30, total 7482L. (Armstrong 106)


Mules are less temperamental and easier to train to carry a load than horses. Can withstand greater environmental extremes and survive on less food an water. Negotiates rough and steep ground better. (Landels 171-2).


Horses are more expensive, more finicky, and more vulnerable to disease and accident than mules, and hence they are more likely to be used as riding horses by the merchants and guards, than as pack animals. (HNC)


Oxen were less subject to disease, they were less likely to be stolen, and they made a good meal if food became more important than transportation. (HNC)


Camels could carry heavy loads further than horses or mules, and required little food and water while traveling. However, when horses encountered camels, they tended to stampede, disrupting wagon and pack trains. ((Huston 157).

Based on the data in section 2, we can reconstruct the cost of transporting your own goods. Even if you don’t want to do this, the analysis is worth conducting, because it acts as a “reality check” on the published cost of transporting by “common carrier”. To arrive at a do-it-yourself cost, we first need the acquisition (purchase or rental) costs of the vehicle (ship, wagon, etc.), service animals (horses, mules, etc.), and crew. Then we need to calculate the operating costs (food, lodging, tolls, etc.) for the route, which will of course be dependent on the length of the route and the speed which we can make upon it.

Riding Post. If you have a light but highly valuable cargo (jewels or legal documents, perhaps), and you are traveling on a major road, you may be able to “ride post,” changing horses every ten miles or so. The 1635 British rate was 2.5d/mile. If you carried ten pounds of cargo (remember, the horse is carrying the rider as well as the cargo, and for speed you want a light load), and made 100 miles a day, that would be an effective long ton mile rate of 560d (2.33 pounds sterling) a day!

In theory, a merchant, or group of merchants, could set up their own post system on any good road. Maintaining one horse at each of a series of inns ten miles apart would carry a cost of about two shillings a day per horse. You would need at least one extra horse, so a post boy could ride through, leading horses back to their home posts. Add another two shillings a day for the post boy. So a de minimis 100 mile post system would cost 24 shillings a day. The long ton mile rate, if it was used by just one courier a day, would be 54 shillings (2.7 pound sterling). For the private system to be cheaper than the British official post, it would have to be used by two couriers a day. But then a courier might be delayed because an earlier courier had taken the post horse to the next post and the post boy had not yet returned it. The solution is to have more horses per post, and more post boys. But that of course increases the system cost.

Since the Royal Post had a habit of seizing private horses when needed to speed a messenger along, some towns kept horses reserved for postal use. Leicester at one point kept six post horses at an annual cost of 33s4p apiece. Later it reduced the lot to four horses, ready at one hour’s notice, for an annual cost of 28s6d each. (Walker, 116) That implies a daily cost, covering horse rental, feed and lodging, of just 1d a horse (rather than the 2s assumed above).

Pack Mule and Pack Horse Trains. A typical long-distance mule train would have one wrangler for every six or seven mules (HNC). If each mule is carrying 300 pounds of freight, then there is about one short ton per wrangler. A mule train which is limiting its journey to a safe region might have twice as many mules per handler, yielding twice the “cargo efficiency”.

While it is possible to rent a horse or mule, that is usually for a jaunt in the local countryside, not for a long-distance journey. Let’s assume that we buy six animals at a price of thirty pounds each, reselling, after a year, at half price. That means that the effective rental price is fifteen pounds a year, or about one shilling a day, each. (Which is consistent with the local rental price.) Food and stabling for each animal is another shilling a day, and wages, food and lodging for the wrangler, another one to three shillings. So a pack train consisting of one wrangler with six packhorses would cost 13-15 shillings a day. If they made 20 miles per day, and each carried 300#, then the cost is 13-15 shillings transporting 0.9 tons 20 miles. The effective cost is then 6.5-8.1 pence per ton mile. But if they only progressed five miles a day, the cost would be 26-32 pence per short ton-mile.

The up-front cost of a minimal pack train is 180 pounds sterling for the animals, one month’s salary (1.5-4.5 pounds) for the wrangler, and expense money enough to cover them until the next major stop (where they could use a letter of credit). For safety’s sake, it is probably desirable to run a longer pack train, so there is more than one wrangler.

Wagon. Wagons are practical for long-distance traffic only where improved roads are available. A wagon with four horses and a driver might be rented for local work for 12.5 shillings a day, and I am going to assume that the amortized purchase price of the ensemble would be around the same number. Add another five shillings for their food and lodging, and, oh, round it up to 18 a day. The Conestoga wagon transports three short tons, and top speed is again twenty miles a day. So the rate per short ton mile is just 3.6 pence, about half that for the pack train. However, if the improved road is a turnpike, the tolls could wipe out the difference. Also, if the turnpike had a regulation limiting the load to just one ton (as England did in the mid-seventeenth century), then even without tolls, the ton mile rate for even a fast-moving wagon would jump to 10.8 pence.

River Barge. If it is just floating downstream, then you just need one man to steer it. If it is being towed by draft animals, then someone needs to be on the towpath, too. A typical crew was three men, even for a barge as large as sixty tons. (Willan 106, cp. Emerson 254).

Ocean-Going Ship. The cost of a ship should be roughly proportional to its surface area, since that will determine the amount of wood (or iron, in the nineteenth century) needed for the hull and deck. On the other hand, the cargo capacity of a ship should be roughly proportional to its volume. Nonetheless, comparative purchase cost figures are usually expressed as so many dollars per long ton. That ton is the displacement ton (35 cubic feet) for warships, and for merchant ships may be the register ton (100 cubic feet) or the actual cargo capacity (usually assumed to be 40 cubic feet per ton).

The next thing we need to consider is the operating cost. Sailing ships provided high cargo efficiency. Even the old cog could carry ten tons of cargo for each sailor. The epitome of the merchantman, the unarmed Dutch Flute (fluyt), had very efficient rigging, and as a result could carry fourteen tons for each crew member operating it. (Kohn I, [get page #]). The average Dutch ship of 1669 was one hundred tons, and had a crew of eight. (Braudel III, 190).

In Elizabethan times, British merchant sailors were paid almost twenty shillings a month (~8d a day), and were also fed at the shipowner’s expense. In 1700, the median common seamen’s monthly wage was 1.23 pounds sterling, or about 25 shillings.

Using British Royal Navy rations (Emerson 158, Pope 150-9), and a variety of sources to price them out, I arrive at a victualing cost of another 8d a day.

So the labor cost (wages plus provisions) is 16d a day, per sailor. If the ship carried ten tons of cargo per sailor, that is 1.6d per ton-day, and if the ship traveled sixty miles in the course of a day, 0.027d per ton mile.

However, this ignores the efficiency losses from waiting for cargo. For example, the Bristol to Bordeaux run, which was about 600 miles, and might take 10 days sailing time, was actually part of a 93 day round trip, when waiting time in Bordeaux was included. So multiply 1.6 by 93, and then divide by 1,200, to obtain its true ton mile labor cost of 0.124d. (If, while in port, half the crew is given shore leave at a time, and must pay for its own food, then cost is reduced to 0.1d.)

In the 1540s, the actual total freight charge in that wine trade, from a Bristol carrier, was about 0.33d per ton mile. In December, 1540, the Jesus of Torres charged only about 0.17d per ton mile.

Unless you are dealing in large quantities of bulk goods, the chances are that you will be hiring a common carrier — a railroad, barge, ship or wagoneer — to transport your goods for you. You will pay a price which will reflect both the intrinsic cost of carriage, and a profit to the carrier.

As explained in my article “Hither and Yon”, a standard method of expressing transportation costs is in the form of the rate per ton mile, which allows comparison of shipments of different tonnages and over different routes. One purpose of this FAQ is to explain how the cost per ton-mile data in that article were derived.

The raw data on costs takes the form, “in year 1XXX, it cost Y (in ducats, guilders, florins, thalers, pesos or whatever) to move goods G from point A to point B.”

If we know the cost for goods which traveled on different routes, then we can put them into the common unit, the cost per ton-mile, by dividing the cost by the number of tons, and the number of miles traveled. If we already have a cost per ton, then just dividing by the distance is enough.

By way of example, it has been reported that in 1807, the cost of road transport between Buffalo and New York City was $100 per ton (Anon. 1918). According to MapQuest, the road distance between these two cities is 400 miles, so that implies a unit cost of 25 cents per ton mile. The opening of the Erie Canal in 1825 reduced the freight charge to $15-25 per ton, and thus to a unit cost of 3.75-6.25 cents per ton-mile (Anon. 1918). By implication, shipment by inland waterway was 4-7 times cheaper than by roadway.

However, please note, when calculating distances, that the true routes are not as the crow flies, but rather must follow the prescribed paths, whether they be roads, canals, rivers or coastlines. The 400 mile route assumed by MapQuest uses roads which might not have existed in 1807. Riparian traffic is even less likely to be direct. For example, those using the Erie Canal first went east to Albany (291 miles), then south on the Hudson River to New York City (156 miles), a total of 447 miles instead of 400. That implies a unit cost of 3.3-5.5 cents per ton-mile, instead of the one previously calculated.

In the case of open water transport, it makes a big difference whether the ship hugs the coast (as was still common in the seventeenth century) or takes the most direct route. As pointed out earlier, maritime routes are often more roundabout than overland ones.


A more difficult problem is how do you find the 1632 equivalent of a cost incurred many years later? In section 5, I will explain how I manipulate data from other time periods. The method is far from perfect, but it allows one to generate plausible equivalents, which is really all that we can hope for.


Sometimes, the history books state the transportation cost as a percentage of the purchase cost. You can back-calculate the absolute transportation cost if you can find the purchase price somewhere; my research so far suggests that this is likely to be available only by checking primary sources. Alternatively, you can just figure out the percentage of purchase cost incurred per travel mile, but then the number can only be used in connection with the same type of goods. Because this data is harder to work with than the absolute costs, I have relegated it to an Appendix.

Here I present the mammoth tables of ton-mile data which I have assembled. In some cases, the ton-mile field is left blank because I have not determined the length of the route in miles. If a reader supplies that information (but bear in mind the warnings above), I can then fill in the ton-mile field.

TimeCost Then1632 Eq (d)Comments
British data
Eliz0.44-0.88d4.8-9.7dscheduled pack trains, Kent to Southampton, 2-4d/cwt. (Emerson 249) Assumed Kent=Maidstone, so ~90 mi. Equiv uses lowest CPI in reign of Henry VIII which was 9 in 1551 (1632=100).
16458.9d- 11.4d9.2-11.8dnear London, army supply during Civil War (Nusbacher)
1692-171014-25d14.8-26.5dBuckinghamshire to London by wagon (Albert 260) Assumed 1700.
170519.2d21.7road (London to York) 16s/cwt for 200 miles (Albert 180)
1710  Northampton to London, 4s/cwt (Albert 180) [need mileage]
17506.3d-8.3d6.6-8.7dturnpike (London-Bristol, 116 mi.)(Albert 173)
175112d12.6turnpike (York to London, 200 mi.)(Albert 174)
175116d16.1turnpike (York to Bridlington or Scarborough, 40 mi.)(Albert 174)
1700-0914.64d15dturnpike, 52 city pairs, Bogart Table 8. Allen CPI avg. 97.75.
1710-1914.76d turnpike, 32 city pairs, Bogart Table 8. Clark (2001) CPI 1.04 (with 1700-09 =1)
1720-2914.76d turnpike, 34 city pairs, Bogart Table 8. Clark (2001) CPI 1.04 (with 1700-09 =1)
1730-3915d turnpike, 45 city pairs, Bogart Table 8. Clark (2001) CPI 0.95 (with 1700-09 =1)
1740-4915.12d turnpike, 58 city pairs, Bogart Table 8. Clark (2001) CPI 0.95 (with 1700-09 =1)
1750-5914.88d turnpike, 52 city pairs, Bogart Table 8. Clark (2001) CPI 1.06 (with 1700-09 =1)
1774  3s-4s per cwt, Birmingham to London. (Gregory 179). [need mileage]
Late 18th16d turnpike (London to Devon 1758-68 or Cornwall 1765-6)(Albert 173)
1840s?6d England, coal, 6d per ton mile (White)
Early American data
1770s  Philadelphia or Baltimore to Pittsburgh, 45s /cwt (Meyer 78)
late 1700s  Philadelphia to Uniontown, $5/cwt (Meyer 78)
1789  Philadelphia to Carlisle, packhorse, 30s ($4) /cwt (Meyer 78)
late 1700s  wagon, Hagerstown to Brownville, over Virginia or Braddock Road, one month trip, $3/cwt (Meyer 78)
1790  Schenectady to Rome (3 men for 9 days), $14 a ton (Meyer 79)
1792  Genesee to Philadelphia, wheat, 1s a bushel (60 pounds)
17941.1s9.8Middletown to Philadelphia, 100 miles, 5L10s a ton. To transport 20 tons, need 20 men and 80 horses. (Meyer 79)
Later American Data (Cost then in cents per short ton mile)
 then cents1632 pence 
1750-185014-183.4-4.3(Seidler, quoting Rothenberg, 1992) Conversion assumes 1$ 1800 =2s 1632.
1800-18203-7 (Seidler, quoting Taylor, 1951)(note the disagreement!)
1790s13 turnpike road (Meyer 208)
1820s?47 dirt road, Baltimore to Wheeling, $125/T, 265 mi., five weeks transit time (Dilts, 283, 292).
E19C20-255.6-7ordinary road (Rae 21) One 1825$=2.33s 1632
E19C15 turnpike (Rae 21)
~18004210turnpike, Pittsburgh to Philadelphia, $5/cwt so $100/ton. (Rae 22).
1808  measured goods, Philadelphia to Baltimore, 12 cents per cubic foot. (Meyer 81).
181340 Army allowance to officers for baggage transport, $2 per cwt per 100 miles. (Meyer 92)
181762.515.8baggage, Philadelphia to Pittsburgh (240 miles), $7.50/cwt. (Atherton, 82)
1818  wagon, NY to Detroit, $4.50/cwt so $90/T. (Meyer 84)
183582.7freight, Philadelphia to Pittsburgh (240 miles), $1/cwt or less. (Atherton, 83)
1830s14-21 Conestoga wagon, Baltimore to Mt. Vernon, 397 mi., $4.25/cwt going, $2.75/cwt return. (Shumway 73)
1830s18 turnpike (National Road), Baltimore to Wheeling, $48/T, 265 mi. (Dilts, 283)
1840s13 same turnpike but now $35/ton, two weeks transit time (Dilts, 283)
1840s8 turnpike, Cumberland to Wheeling 130 miles, $10/ton, 6 day transit (Dilts, 292)
1846  wagons on Santa Fe Trail, Fort Leavenworth KS to Santa Fe NM, $11.75/cwt. (Settle 30)
1850  Id., $13.47-14.33/cwt (Settle 37)
185523-445.9-11.3Conestoga wagons, $1.14-2.20 to move 1 cwt military stores 100 miles, April or May loadings, Kansas to New Mexico to Utah (Settle 42)
185722-705.9-18.7Conestoga wagons, New Mexico route, $1.12-$3.50 to move 1 cwt 100 miles. (Settle 51)
185725-716.7-19Conestoga wagons, Utah route, up to 1,200 miles, $1.25-$3.55 to move 1 cwt 100 miles, depending on season. (Settle 51)
1860>=15 Atack
Mid 19th15 Anon. 1918
1895254wagon (Rae 33)
1946<3 US, trucks 10 tons freight 300 mpd so 3000 ton miles, driver paid $16/day so labor cost is 0.5 cents per ton mile. Rest of cost is fuel, tire wear, depreciation, etc. Turkey, carts 0.25 tons 5 mpd so 1.25 ton miles/day. Driver is paid $1.25/day so labor cost is $1/Tmi. (Rae 109-110)
1860s10014.4-28.8human packers, $5 to carry 100# for 100 miles, Hazelton to Fort Babine. (Skeena)

1s=12d. 20 cwt=1 ton.

Bogart’s data is from the Middlesex, while Albert’s is from Middlesex and Westminster.

American data (Cost Then in cents per short ton mile. 1632 in pence.)
YearCost then1632 
18502-90.66-3.0(based on Cincinnati to NY) (Merchant)
18602.060.59(New York Central)(Atack)
Mid 19th1.5 Anon. 1918
19180.8 Anon. 1918
18313.531.02horse railroad. (Meyer 588)
E19C4 trains at 4 mph (Dilts 41)
1840s1.50.44Baltimore to Cumberland, 395 miles, $6/ton, 30 hours, 13 mph (Dilts 292). 1US$=2.42 1632 sh.
184220.59coal, Baltimore & Ohio RR. (Dilts 299). 1.33 cents on new branch line. (301). 1.75 (311).
18430.7150.23Philadelphia & Reading, for merchandise. For coal, 0.484. (Poor, 485) Poor also gives ton mile data for 1844-1859.
1840s4-8 flour (Dilts 301, 308)
1820s0.5-1 Pa RR. 0.5 if freight accompanying passenger, 1 otherwise. (Gordon 19-20)
18485.3-7.51.6-2.2Detroit Central. Coal, iron, grain, 39 cents/cwt, Dry goods, 55 cents/cwt, Detroit to Kalamazoo, 146 miles. (Meyer 505)
18481st: 8.97 2nd: 6.16 Avg 7.561st: 2.65 2nd: 1.82 Avg: 2.24average freight rates (cents) per ton mile per Daggett, Railroad Guide: (First class then second class) ME: 5.68, 3.38 (226.75 mi) NH: 5.25, 5.00 (99) VT: 4.00. 4.00 (33) MA: 5.47, 4.54 (1929.5) RI: 6.37, 4.39 (91.25) CT: 5.75, 3.50 (253.75) NY: 9.04, 5.79 (798) NJ: 13.57, 11.66 (155) PA: 6.75, 5.25 (355) MD: 4.56, 3.12 (661) VA: 10.44. 4.69 (264.5) NC: 9.83, 6.37 (248) SC: 10.75, 5.50 (204) GA: 9.33, 4.78 (602) KY: 9.00, 9.00 (28) MS: 24.39, 17.30 (70) AL: 16.83, 8.00 (67) OH: 6.60, 4.62 (307) IN: 8.00, 5.81 (86) MI: 8.44, 6.50 (241) final avg. is for 117 RRs with total length of 6,720 miles. (Meyer 555, 576)
18513.1-5.2 Georgia Central. 30-50 cents/cwt depending on class, Savannah to Macon, 191 miles. Or, for measurement goods, 13 cents/cu.ft. (Meyer 481). implies 4 cu ft = 1 cwt or 80 cu ft = 1 measurement ton.
18515-8.81.7-2.9Macon and Western. 25-44 cents/cwt depending on class, Macon to Atlanta, 101 miles. Or 9 cents/cu.ft. (Meyer 481)
18513-8 Western and Atlantic. 15-40 cents/cwt, Atlanta to Dalton, 100 miles. Or 8 cents//cu.ft. (Meyer 481)
18513.2-7.4 Georgia RR. 65 cents-$1.50/cwt, Charleston to Dalton, 408 miles. Or 28 cents/cu.ft. (Meyer 481)
18542-3.5 NY to points west 2.4-3.1 cents/ton mile Buffalo to Chicago 2.5 cents/ton mile Boston to Great Lakes, 2.0-3.5 cents/TM (Meyer 581)
1857? NY RRs. NY to Cincinnati, $0.70-$1.80, depending on class, per cwt. NY to Indianapolis, $0.75-1.50/cwt. (Meyer 570)
18590.511 Philadelphia and Reading. For coal, 0.242 cents per ton mile. (Poor 485)
1854-592.8 New York Central. Total ton miles freight, 783,069,838-10,861,290= 772,208,548. (Poor 277). Gross earnings freight, 21,947,162 (Poor 278).
1852-592.3 New York and Erie. Total ton miles freight, 1,193,062,944. (Poor 286). Gross earnings freight, 1883198 + 2537215 + 3369590 + 3653002 + 4545722 + 4097610 + 3843310 + 3195869 =$27,125,516. (Poor 286)
1855-592.5 Pennsylvania RR. Total ton miles freight, 672,462,873. (Poor 473). Total gross earnings freight, 2805306+ 3244291+ 3374041+ 3536206+ 3656111= $16,615,955. (Poor 474).
18591.7 Pittsburgh, Fort Wayne and Chicago RR. Total ton miles freight, 58,421,205. (Poor 497). Gross earnings freight, 964,133. (Poor 498).
European data, Ocean Trade
1539-410.4d1.19dwine from Bordeaux to Bristol, Bristol carrier, sailing time ~10 days one way. Average round trip 97 days, including waiting time. 20s/ 600 miles. “Foreign” carriers charged 13-20s. (Jones, I, 15)
15400.27d1.26dwine from Bordeaux to London. This was an upper limit fixed by statute. 18s / 800 miles. (Jones, I, 16)
1539-410.22d wine from Southern Iberia to Bristol, Bristol carrier, sailing time ~20 days one way. Avg. round trip 153 days, including waiting time. 25s/1350 mi. (Cadiz) “Foreign” carriers charged 15-25s. (Id.)
15400.18d0.57dwine from Seville to London. This was an upper limit fixed by statute. 23s/1500 mi. (Jones, I, 16)
1695? 1.5L for one moio (24 bushels, 814 liters) wheat from Azores to Madeira (Duncan 68-69)(cost doubles or more in winter)
16960.125d0.11d15s for one pipe (476.94 liters) of wine, Madeira to British West Indies (Duncan 50) 2991 naut. mi. (3440 land mi) Funchal to Santo Domingo. 28.32 liters = 1 cu ft. So 0.42 “measurement tons”.
1650? 50,000L for 100 pipes (47,694 liters) wine from Funchal to Bahia (cost is only 25% if using own ship)(Duncan 49)
18490.05d clipper “Oriental”, 1003 tons. For NY to Hong Kong, std rate charged 3L10s per long ton (40 cu. ft.). Trip was 81d, avg. 200 mpd, i.e., 16,200 mi. 3L10s=70s=840d. Chartered rate was 6L per long ton. (Armstrong 93)
American data, Ocean Trade
~18000.25 cents0.06dEurope to USA. $9 (or 30s) for one ton for 3600 miles (not 3000 as stated in ref). (Meyer 81, 90). Also gives measured goods price of 20 cents per cubic foot, implying that one ton is 45 cubic feet.
British Data, Inland Waterways
16050.65d0.95dcoal, Nottingham to Gainsborough [~55 miles] on River Trent (flat rate of 3s/T (Willan 120)
16050.48d0.7dcoal, Gainsborough to Hull [~50 miles] on River Trent (flat rate of 2s/T) (Willan 120)
16150.8d0.89dglass, Nottingham to Hull [~105 miles] 7s/T (Willan 120)
16366.67d6.82dMildenhall to Bury St. Edmunds on River Lark ~12 miles, for 6s8d (Willan 27) Mildenhall to Worlington, for 2s8d.
1630s1d1.02dLondon to Wallingford (upstream ~ 60 miles measured from London Bridge), coal, 5s/T. But with tolls, total cost=15s/T (3d/T-mi)
16670.36d0.4dKing’s Lynn to Cambridge (~44 mi), coal, 4s/chauldron (chauldron=3 tons)
16960.67d0.59dOuse, King’s Lynn to St. Ives, coal, 3s (Willan 120)
17291d1.06dcoal (Willan 119)
17207-9.5d7.35-10dwater, East Retford to Bawtry (winter-summer)(Willan 121)
American data (shillings per ton mile)
1770s  flour, Pittsburgh to Pensacola via Ohio and Mississippi Rivers, one month travel time, 25s a ton (Meyer 78)
1780s  flour, Philadelphia to Pensacola, 5s/ton. (Meyer 101)
17940.34s3dMiddletown to Philadelphia by canal (132 miles), 2L5s a ton (Meyer 78). For 20 tons, total is 45L, of which 40-6-3 is tolls and rest is hire of man, boy, horse and boat for 5 days.
18170.45s2.6dfloat, down Ohio River, 2s3d/cwt -100 miles (Meyer 87)
American data (cents per short ton mile)
late 1700s  flatboat and barge, St. Louis to New Orleans, $6.75/cwt (Meyer 78) Reduced over 50% by steamboat, 1816 (Meyer 103)
   Pittsburgh to New Orleans. $6.75/cwt general merchandise. Decreased >50% by steamboat in 1811. (Meyer 103)
1796  salt, New Orleans to Nashville, 37.5 cents a bushel (Meyer 79)(bushel of coarse salt is 70 pounds, of fine salt, 50).
19C1.50.45d @2.5s /$Canals (Dilts 51)
1825  $5/ton Buffalo to NY (Dilts 25)
1831  steamboat, Pittsburgh to Lexington, KY., $1.50/cwt. (Atherton 89)
18344.5, 2.1, incl. tolls1.5d, 0.7dErie Canal, 363 miles, one ton Albany to Buffalo, $6.57 tolls and $9.83 cost of carriage, for total of $16.40. In opposite direction, $3.28 tolls and $4.40 carriage, total $7.68. (Poor 364)
18590.7, 0.8, incl. tolls0.21d 0.24dErie Canal, 350.5 miles [note the enlarged canal constructed in 1837 was shorter than the old one], one ton Albany to Buffalo, 70 cents tolls and $1.70 carriage, total $2.40. In opposite direction, $1.41 tolls and $1.46 freight, total $2.87. (Poor 364) Poor also gives totals for 1835-1858.
18540.5-2.40.13d to 0.63dHudson River, 0.7 cents per TM Erie Canal, 1.1 Great Lakes, 0.5-1 Pa. Canal 2.4 Ohio R 0.8 Miss. R, lower, 0.6 Miss. R, upper, 0.9 Ohio Canal 1 Wabash & Erie Canal 1.9 Illinois Canal 1.4 Illinois River 1.2 (Meyer 581)
18600.990.29dcanal (Erie)(Atack)
18600.370.11driver (Mississippi)(Atack)
18600.100.03dopen water (Great Lakes)(Atack)

[is 1860 data steamship-based?]

1807?$120$28Newcastle to Frenchtown, 50 tons, 25 wagons by land or one barge by canal. Canal cost is $3 for actual carriage and $25 in tolls. (Meyer 81)
1810$1$0.75One ton Utica to Schenectady. (Meyer 82)
1817?$90 One ton Albany to Buffalo (Meyer 86)
1830-50 7.78” (Meyer 86)
183915-20(T) 10-15(M) 2.5 (RR)1.5(C) 2-4(L) 0.5-1.5(R)estimated cents per ton mile, by Charles Ellet Jr. (chief engineer on James River and Kanawha Canal and Railroad) (T) common turnpike (M) macadam road (RR) railroad (C) canals, irrespective of tolls (L) steamboats on lakes (R) steamboats on Ohio and Mississippi R’s (Meyer 574)
1810?$75-100/T16-32/TSeneca Lake to Albany (Meyer 174)
182513 cents /TM0.4 cents/TMPa. Land: 6 horse+ 1man carry 2 tons in wagon, 20 mpd. 6 horses @ $0.75/day=$4.50, 1 man @ 0.75/day, total $5.25. Canal also 20 mpd, but boat carries 25 tons, needs 1 horse, 1 man, 1 boy. Total $2. (Meyer 225) Ignores cost of wagon and boat.
19C  road, Vicksburg to Jackson, 46 miles. $1.50-2.00/cwt. (65-87 cents/ton mile). RR in 1851, 30 cents/cwt. (13 cents/ton mile). (Meyer 475).
18175s1s8done cwt Louisville to Peoria (down), 307 miles (Meyer 87)
181720s3d4s3done cwt Peoria to New Orleans (down), 1,130 miles (Meyer 87)
181818/0-22/63/4.5-4/6one cwt, Louisville to New Orleans, 1,412 miles (Meyer 87)
1810?$60-75$30One ton, Buffalo to Montreal (Meyer 168 n.3)

In 1801-7, the upstream commerce of the Miss. R. was only 10% of that downstream. (Meyer 109)

Table 4.2.6 Relative Cost data

Thirteenth Century

1300: Sicily to Northern Italy, by sea, grain

+40% (Kohn I p. 37, 52)

Sicily to Northern Italy, by sea, cotton

+7% (Kohn I, p. 52)

Gascony to England, by sea, wine

+10% (Kohn I, p. 52)

Bordeaux to British Isles or Low Countries, by sea, wine

+10% (Kohn I, p. 37)

Fourteenth Century

(early) Bruges to Florence overland, “Caen says” (a relatively inexpensive woolen)

+8% to cost. (Kohn I, p. 51)

1321: Armenia to southern Italy, by sea(?), grain

+160% (Kohn II, p. 16)

England, overland, grain

+0.4% per mile (Kohn I, p. 51)

1398: Wervik, Belgium to Barcelona, Spain by land (Expedia says that the present shortest driving distance is 758 miles), woolens, +22%; by sea, woolens, +15% (Kohn I, p. 53)

Asia to Europe, overland, silks and spices

+8% (Kohn II)

Fifteenth Century

~1400, Genoa to England, by sea, alum

+7% (typical profit on alum was over 25%)

~1400, England to Genoa, by sea, wool

+4% (typical profit on wool was over 35%)

Baltic to Low Countries, by sea, grain

+100% (Kohn I, p. 37)

Portugal to Bruges, by sea, salt

+600% (Kohn I, p. 37)

Sixteenth Century

Sicily to Italy, by land then by sea, grain

+125% (if land trip avoided, +95%)(Kohn I, p. 52, citing Parry, 1967)

England, overland, coal

+10% per mile(!) (Kohn I, p. 51)

1584: Spain to Tuscany, grain

+120% (Kohn I, p. 16)

1590, Poland to Venice, overland, grain

+300% (Kohn I, p. 52)

Sicily to Spain, by water, grain

+40% (Kohn I, p. 52, citing Braudel (1972) p 578)(can be reconciled with the earlier figure if we ignore the export license and the insurance)

Ancona to Rome, overland (100 miles), grain

+80% (Kohn II, p. 15)

Sicily to Rome, by water (500 miles), grain

+60% (Kohn II, p. 15)

China to Middle East (one year journey), overland, silk and spices

+14% (Kohn II, 34, citing Rossabi, 1990).

Turkestan to Kabul, overland, goods not stated, but clearly luxury goods since the final price was 5x cost

+100-200%(Samhaber 96).


The nominal prices of goods and services change over time. If we have a ton-mile rate for one time period, and need to convert it to a different one, we need a conversion factor.

Turner has published (on a very useful article on the European currencies of 1632. Turner only briefly touches upon the issue of changes in purchasing power with time. The buying power, in present money (year unstated), of one guilder in 1632 is said to be 24 modern British pounds (pre-Euro, I assume) or US$36 (implying that one 1632 shilling is NUS$18).

Using the IISG’s British inflation calculator, two English shillings in the year 1632 is equivalent to 10.49 pounds in the year 2000 (this calculation is based on the retail price index). During 2000, one British pound was roughly equivalent to US$1.50, implying that two 1632 shillings was equivalent to only about US$16 at the time of the Ring of Fire. The Grantville Gazette story “Other People’s Money” posits an exchange of one guilder to US$42, implying that the 1632 shilling is worth $21.

Likewise, according to the IISG’s Dutch inflation calculator, one Dutch guilder (florin) in 1632 is equivalent to 19.54 guilders (florins) in 2000. (Note that the rate of devaluation is different for English and Dutch currencies.) The average 2000 exchange rate into dollars was about 0.4, yielding an equivalent of only US$8.

Given the dominance of labor costs on costs of carriage, it arguably is best to use a wage index, especially one for unskilled laborers. The other approach is to use a consumer price index, which is essentially the cost of a “basket of goods”.

Allen, Robert, “Wages, Prices and Living Standards: The World – Historical Perspective”, has published online spreadsheets (Excel files) setting forth the daily wages of building laborers and building craftsmen, with currency payments expressed as grams of silver, for Antwerp, Amsterdam, London, Oxford, Paris, Strasbourg, Florence, Milan, Naples, Valencia, Madrid, Augsburg, Leipzig, Vienna, Gdansk, Krakow, and Warsaw, beginning as early as 1264.

Both Excel files also include Allen’s Consumer Price Indexes for these cities. The methodology of computing the CPI is described in Allen, “The Great Divergence in European Wages and Prices from the Middle Ages to the First World War,”

My original inclination had been to use Allen’s laborer’s wage data. However, it is relatively spotty. For example, the same number is given for London for 1620-1639 (7.423871) and again for 1640-1649 (8.351855). Hence, I decided to use Allen’s CPI, which, being based on more than one “commodity”, is more responsive to year-to-year economic changes.

The following is a simplified British CPI, derived from Allen’s, which rounds the annual data to the nearest 1.5-fold change. (CPI for London in 1632 set at 100.)

1500: 30

1501-1502: 44

1503-1518: 30 (except for 1512, CPI 44)

1519-1521: 44

1522-1534: 30 (except for 1527, CPI 44)

1535-1536: 44

1537-1544: 30 (except for 1541, CPI 44)

1545: 20

1546-1548: 13

1549-1550: 20

1551: 9

1552-1554: 44

1555-1556: 67

1557-1575: 44 (except for 1562, 1573; CPI 67)

1576-1577: 67

1578-1584: 44

1585-1586: 67

1587-1588: 44

1589-1593: 67 (except for 1591, CPI 44)

1594-1597: 100

1598-1606: 67

1607-1608: 100

1609-1610: 67

1611-1617: 100

1618-1620: 67

1621-1625: 100

1626-1627: 67

1628-1646: 100

1647-1649: 150

1650-1771: 100 (except for 1661, 1693, 1709, 1767; CPI 150)

1772-1778: 150 (except for 1776, CPI 100)

1779-1780: 100

1781-1783: 150

1784-1789: 100

1790-1799: 150 (except for 1792, CPI 100)

1800-1801: 225

1802-1804: 150

1805-1820: 225 (except for 1815, CPI 150)

1821-1854: 150 (except for 1847, CPI 225)

1855-1856: 225

1857-1866: 150

1867-1868: 225

1869-1871: 150

1872-1893: 225

1894-1913: 337

The colonial American data, and even some early nineteenth century data, is expressed in English currency. I have assumed that there is a one-to-one relationship between the value of British shilling in America, and its value in Britain the same year. However, in fact it is likely that the British shillings traded here at a premium price.

Most of the ton-mile data that is readily available is for nineteenth century America. There are two considerations with the use of this data. First, for all post-colonial American figures, it is safe to assume that the ton is the U.S. short ton, 2000 pounds, not the British long ton, 2240 pounds.

Secondly, it will be necessary to convert the nineteenth century U.S. dollar to a 1632 equivalent. Plainly, there is no American consumer price index or dollar-pound exchange rate for 1632! Hence, we don’t have the choice of first deflating the 1800 dollar to its value in 1632, and then converting it at 1632 exchange rates to the British shilling (or Dutch guilder). Instead, I converted dollars to pounds using a contemporary exchange rate (EHS), then deflated the nineteenth century British amount using the Allen CPI for London.

The American/British exchange rate fluctuates daily. In 1999, one British pound bought anywhere from $1.56-1.66, depending on the purchase date. And the dollar and the pound are among the world’s more stable currencies. The EHS exchange rates are supposedly average values. Note that for the periods 1817-1861 and 1877-1918, the annual average was in the range of one U.S. dollar equals 0.20-0.22 British pounds. For 1866-1876, the equivalent was 0.15-0.18 British pounds. Over the longer period 1791-1918, the highest equivalent was 0.28 (1812), and the lowest 0.1 (1864).

Combining the EHS exchange rate data with the Allen CPI, we find that the equivalent of one U.S. dollar in 1632 British shillings is as follows:

1791-1799: 2.8-3.6

1800-1862: 2.0-2.9

1863-1869: 1.2-1.7

1870-1893: 1.6-2.2

1894-1913: 1.0-1.5

There are, of course other European currencies, such as the Dutch guilder, which in theory could be deflated to 1632 using a Dutch CPI and then converted into British shillings, rather than converted first and then deflated.

In general, you cannot assume that if you convert currency A to currency B at the 1632 exchange rate, and then multiply by the year 2000 purchasing power of one unit of currency B, that you will get the same result as if you first multiplied the original amount of currency A by its year 2000 purchasing power, and then converted currency A to currency B at the 2000 exchange rate. The technical term for this is that there is a lack of purchasing power parity between the two currencies.


Units of weight: a British long ton is 2,240 pounds, an American short ton is 2,000 pounds, and a metric ton (tonne) is 1,000 kilograms or 2,204 pounds. There are 20 long hundred weights (each 112 pounds) in a long ton and 20 short hundred weights (each 100 pounds) in a short ton. Unless specifically stated otherwise by the reference, I assume that European data referring to tons and hundredweights means the long (British) form, and that American data means the short (American) form. Any reference to “tonne” is assumed to mean the metric ton. A “stone” is 14 pounds and a “quarter” (of a long hundred weight) is 28 pounds.

A short ton of sea water has a volume at 13oC. of 31.3 cu. ft. A long, or British ton – 2,240 pounds – of sea water occupies 35 cu. ft. The density of wine depends on its alcohol and sugar content. The specific gravity of sea water is around 1.035.

For bushels of various sorts, see  

For a “fanega”, see

British Currency. There are 12 pence (d for denarius) to the shilling (s for solidus), and 20 shilling to the pound (L for libra).


Albert, The Turnpike Road System in England: 1663-1840 (Cambridge U. Press, 1972).

Bogart, “Turnpike Trusts and the Transportation Revolution in Eighteenth Century England,”

Crofts, Packhorse, Waggon and Post: Land Carriage and Communications Under the Tudors and Stuarts (1967).

Gregory, The Story of the Road (1938).

Hindley, A History of Roads (1971).

Nusbacher, Aryeh J.S., “Civil Supply in the Civil War: Supply of Victuals to the New Model Army on the Naseby Campaign”, 1-14 June 1645, Eng. Historical Review 115: 460 (Feb. 2000).

Holbrook, The Old Post Road (1962).

Rae, The Road and the Car in American Life

Shumway, Conestoga Wagon 1750-1850 (1968).

Settle, War Drums and Wagon Wheels (1966).

Tarr, The History of the Carriage (1969).

Walker, Haste, Post-Haste (1938).

Huston, The Sinews of War: Army Logistics 1775-1953 (1966)

Rivers and Canals

Drago, Canal Days in America (1972).

Guignino, “The Farmington Canal 18221847: An Attempt At Internal Improvement” (1981),

Hadfield, World Canals: Inland Navigation Past and Present (1986).

“International Canal Monuments List”,

Poor, History of the Railroads and Canals of the United States of America, Vol. I (1860; reprinted 1970)(great for cost statistics).

Willan, River Navigation in England, 1600-1750 (Frank Cass & Co., Ltd.: 1964).

Drago, Canal Days in America (1972).

Sailing Barge Ass’n, “Why the Thames Sailing Barge is so Important”,

Coastal and Ocean Trade

Gardner, East India Company

Marshall, Ocean Traders: From the Portuguese Discoveries to the Present Day (1990).

Chapter 1, “The economics of shipping in midsixteenth century Bristol,” and Chapter 3, “The commercial strategies of Bristol’s shipowners, 153943,” in Jones, Evan, The Bristol Shipping Industry in the Sixteenth Century (Ph.D., Edinburgh 1998), online at

Milton, Nathaniel’s Nutmeg (1999).

Duncan, Atlantic Islands

Armstrong, The Merchantmen (1969)

Pope, Life in Nelson’s Navy

Rediker, Between the Devil and the Deep Blue Sea: Merchant Seamen, Pirates and the Anglo-American Maritime World 1700-1750 (1987).

Saville, Hornblower’s Ships: Their History & Their Models (2000).


Dilts, The Great Road: The Building of the Baltimore & Ohio, the Nation’s First Railroad, 1825-1853 (Stanford U. Press: 1993).

Gordon, Passage to Union: How the Railroads Transformed American Life, 1829-1929 (Ivan R. Dee, Chicago: 1996).

Hastings, Railroads: An International History (Praeger: 1972).

Poor, supra.

++Boyd, The American Freight Train

Animal Power

Anon., “Remount in Italy,” Quartermaster Review (Mar./Apr. 19460,

Warfield, “The Quartermaster’s Department, 1861-1864,” Quartermaster Review (Sept.-Oct., 1928),

Historical Novelists Center (HNC), “Equine Data: Speeds, Load Carrying, Feed, and So Forth” (1997).

Thrapp, “The Mules of Mars”, The Quartermaster Review (May-June 1946) online at .

Landels, Engineering in the Ancient World (1978)

[Geocities] “Appendix B: Army Animals”,

Harwood, “Bred to go where horses fear to tread, goats pack quite a load,” Seattle Post-Intelligencer (March 27, 1997),

[Skeena] “Early Transportation on the Skeena River: The Pack Trains,”

Table 4, Hicks, Alastair, “Power and Food Security,” SD Dimensions (October 1997), paper presented at the Int’l Solary Energy Soc’y 1997 Solar World Congress, Aug. 24-30, 1997. Online at ; citing Tools for Agriculture, 1992.

Watkins, “Comparison of Power Produced by Various Draft Animals and Humans,

Thompson, Sled Dog Training Manual,

Savory, “The Mule,” Scientific American (December 1970).

European Prices, Wages, Currencies and Exchange Rates

Anon., “Seventeenth Century Prices and Wages”,

Anon., “Eighteenth-Century Currencies and Exchange Rates”,

Allen, Robert, “Wages, Prices and Living Standards: The World – Historical Perspective”,

Allen, “The Great Divergence in European Wages and Prices from the Middle Ages to the First World War,” )

Fischer, The Great Wave: Price Revolutions and the Rhythm of History (Oxford Univ.: 1996).

Library, British House of Commons, Research Paper 99/20, “Inflation: The Value of the Pound 1750-1998,” (Feb. 23, 1999),

(this overlaps with the date range of Allen’s CPI for London, 1284-1913)

O’Donaghue, Goulding and Allen, “Consumer Price Inflation Since 1750,” (March 2004), (British price data)

Turner, “Money and Exchange Rates”, RTF file at

“The Value of Money”,

International Institute of Social History, “Prices and Wages” . There are three useful links on this page. The IISH’s own data is accessed through the “List of Datafiles”. (One dataset is of prices and wages in the western Netherlands, 1450-1800, and another is of wages and cost of living in London, 1450-1700.) Links to external web sites are collected on the “Index to Other Websites”. And there is access to the historical “Value of the Guilder”, and, indirectly, to the values of the U.S. dollar and British pound.

Nineteenth and Twentieth Century Exchange Rates

Economic History Services, “What Was the Exchange Rate Then?”

(includes exchange rates between US dollar and British pound for 1791-2000)

Ozforex, “Yearly Average Rate Calculator”,

U.S. Prices

Atherton, The Frontier Merchant in Mid-America

Sahr, “Inflation Conversion Factors for Dollars 1665 to Estimated 2014,”

“Daily Life” Studies

Emerson, The Writer’s Guide to Everyday Life in Renaissance England from 1485-1649 (1996).

Singman, Daily Life in Elizabethan England (Greenwood Press: 1995).

Transportation History and Economics

Anon., “Influence of Transportation Upon American Industrial Development” (originally published 1918), online at

[Anon. 1918]

Kohn, Meir, Working Paper, “The Cost of Transportation in Pre-Industrial Europe” (January 2001), (Kohn I)

Kohn, Meir, Working Paper, “Trading Costs and the Pattern of Trade in Pre-Industrial Europe” (Kohn II)

Kohn, Meir, Working Paper, “Trading Costs, the Expansion of Trade and Economic Growth in Pre-Industrial Europe” (Kohn III)

Meyer, History of Transportation in the United States Before 1860 (Carnegie Institution: 1948).

Parry, “Transports and Trade Routes”, Chap. 3 in Cambridge Economic History of Europe (1967).

Ringwalt, “Development of Early Transportation Systems in the United States”

General Economic History


Atack and Passell, A New Economic View of American History, Table 6.3 (2d ed. 1994)

Braudel, Capitalism and Material Life 1400-1800 (Harper & Row: 1967).

Braudel, Civilization and Capitalism, 15th-18th Century, Vol. 1, The Wheels of Commerce, Vol. 2, The Structures of Everyday Life, Vol. 3, The Perspective of the World (U. Calif. Press: 1982-4).

Cipolla, Before the Industrial Revolution: European Society and Economy 1000-1700 (3d. Ed., W.W. Norton & Co.: 1993).

Landes, The Wealth and Poverty of Nations

Merchant, Carolyn, Course ESPM160 Outline, “Farms and Cities in the Early Republic”,

Seidler, “Economic Activity and Changes in Forest Cover: Thoughts on New England”,

Samhaber, Merchants Make History (1960)

Tuma, European Economic History (John Day: 1971).

White and Southall, Electronic Atlas of Industrialising Britain, Chap. 8,


sailing distances between world ports

“Jim Shead’s Waterways Information”

These are unedited raw notes which have not been integrated into the rest of the FAQ but are listed for what value they might have.

“Catalan records of shipping costs show that the cost of arming merchant ships increased freight rates by 25% or more between 1275 and 1330.” (Kohn I, p. 17; Kohn III, p. 22).

Cost of loading and unloading 50 tons, 1808, USA: $10.00 (Meyer 81)

50 ton canal boat required one man and 2 horses. (Meyer 81)

$60mil expended for transportation during War of 1812 (Meyer 91).

Cost of Passenger Train Operation, 1850s

Locomotive power, 20 cents/mile

Passenger car (60 seats), 2 cents/mile

Baggage car, 2 cents/mile

conductor $2/day

brakeman, $1/day

by implication, train goes 100 mpd

Can increase # of passenger cars to four without increase in locomotive power or baggage cars.

engine draws 100 tons

14 passengers with baggage (incl. Share of car) = 1 ton

4 cars, with 240 passengers = 17 tons

Passenger fare is 2.5.. 2.. 1 .. 1 cent/mile for 1-3 cars.

(Meyer 581-2)

Cost of operation of horse railroad 1831

3.53 cents/ton mile

breakdown: mules and horses 1.33, hands, 1.33, repairing wagons 0.66, oil for wagons, 0.2. Ignores depreciation and replacement. (Meyer 588)

current on Miss. R carries boat 80-100 mpd.

Ohio R (high) 3-4 mph.

Can navigate at night on Ohio but not Miss.

(Meyer 115)

State road from Charleston to Columbia SC failed because tolls too high, wagoners took free roads. (Meyer 258).

Cost to portage Louisville falls was 75 cents/ton. (Meyer 280)

RR. First flanged outer edge of rail, then inner edge of rail, then wheels. (Meyer 307) First horse railway, 1807, Beacon Hill (Meyer 308). Roads sometimes had sharp curves or steep grades to keep construction cost down. (Meyer 310). Solid granite track on Boston and Lowell had long life but destroyed rolling stock (Meyer 311). Strap iron rail on Ithaca and Oswego curled and pierced floors of cars. (Meyer 311). Building track on pilings instead of graded road bed was disastrous (Meyer 369).

Lockport and Niagzara Falls RR charges 25 cents/cwt normally, but only 15.5/cwt for carload lots. 24.3 miles. (Meyer 377)

Copper sheathing of ship bottoms began in 1761, improved speed by reducing barnacle drag factor. (Armstrong 41).

American ships at Canton averaged 297 tons (Armstrong 47).

Insurance for ships

1782 20% without convoy, 12% with. (Armstrong 55)

1806 8%, 1812 5.3% (under convoy system)(Armstrong 48)

1860 average tonnage, sailing ship 160 tons, steamship, 220 tons. (Armstrong 118).

Largest sailing ship, France II, 5633 tons, 1911. (Armstrong 121).

Madrid postmaster bribed 28 ducats a month to let Enlgish envoy see all letters from Englishmen. (Walker, 54).

vehicle weights:

England 1621 – four wheeled wagon had maximum weight by law of one ton. 1662 – maximum cargo of 20 cwt (one ton) in winter and 30 cwt in summer. (Albert 133).

For ships, size limit for wood ships was around 2000 tons. (Braudel SEL 423).


Water Travel: A quantum leap in carrying capacity was achieved when you left dry land. The maximum size of the river boat actually depends on the river conditions. On most British rivers, the barges were at most twenty to forty tons (Willan 97), but on the Thames, barges eventually reached a size of 250 tons (Sailing Barge Ass’n). On the Elbe-Lubeck canal, the standard size boat is 19 meters by 3.25 meters, and carries 12.5 tons. (Hadfield 33). The largest Russian barges were 150-170 tons (Hadfield 56).

Large ships were usually built, to order, for a group of merchants, each thereby acquiring a share in the ship. (Kohn I, p. 29).

The typical size of a sailing ship depended on circumstances. In the seventeenth century, the typical English ship trading to Spain was twenty to forty tons. (Kohn I, p. 25). [this seems small to me] Small ships could use smaller harbors, could find full cargoes faster, and could be loaded or unloaded quickly. For the long-distance trade to the Spice Islands, larger vessels, in the 500-1,000 ton range, were used.