This article is part of our Stacked Series.
In the summer of 1904, the Toronto branch of the Society for the Prevention of Cruelty to Animals organized a parade through the city’s main streets. Five hundred horses took part, and they were judged by category, much as any livestock show would judge them today: cartage company horses, brewery horses, coal company horses, the horses of dry goods and clothing stores, hatters and furriers, bakers and confectioners, milk companies, laundry companies, grocers, florists, butchers, and a final catch-all class for any trade the organizers had not anticipated.
That parade is, in effect, a trade-sector inventory of last leg logistics, taking place a full century before the term existed. Every one of those categories was solving the same problem distribution networks still spend enormous resources solving today: how to move goods the final distance from a fixed point of supply to a dispersed set of destinations. The technology has changed almost completely since 1904. The underlying problem has not changed at all.
This is a short history of how that problem has been solved, and re-solved, across roughly a century and a half.
The Horse as Industrial Infrastructure
It is easy, from a present-day vantage point, to think of horse-drawn delivery as a quaint precursor to “real” logistics. The historical record does not support that view. By the late nineteenth century, the horse had become a deliberately engineered component of urban industrial infrastructure, and the scale of that infrastructure was substantial. Demand for horsepower in North American and British cities grew faster than the human population itself through the final decades of the century. In Toronto, the urban horse population rose from 1,874 in 1871 to 7,269 in 1891. By 1900, the ratio of horses to humans in major Canadian cities ranged from roughly one horse per eighteen people in Ottawa to one per forty-one in Saint John, and in American cities, the ratio ranged from one per 7.4 people in Kansas City to one per 26.4 in New York. Historians estimate the world population of working horses peaked between 1910 and 1920 at around 110 million animals, double what it had been a century earlier.
A significant share of this growth was driven directly by the railway. As historians Joanna Dean and Lucas Wilson have argued, far from displacing horses, the expansion of rail networks created substantial new demand for horse-drawn haulage, because the railway solved the long-distance movement of goods but created, at every station, a new last leg problem. Fixed rail lines delivered freight to a depot; horse-drawn carts and wagons provided the flexible linkage between that fixed point and the many individual destinations the freight still needed to reach. Every mile of new track laid in the nineteenth century generated a corresponding need for horse-powered distribution at its endpoints, which is part of why urban horse populations climbed in close step with railway expansion rather than declining in its wake.
The animals themselves were bred and developed specifically for this industrial purpose. Heavy draft horses roughly doubled in average size over the nineteenth century, growing from 900 to 1,100 pounds at mid-century to 1,800 to 2,000 pounds by 1880, as breeders imported heavier European stock such as the Clydesdale, the Shire, and the Percheron specifically to meet urban hauling demand. The payoff in capacity was considerable: a horse could pull roughly five-eighths of a ton on a soft dirt road, but as much as eight to ten tons on a hard rail or tramway surface, which is a meaningful part of why nineteenth-century cities invested so heavily in road paving and why municipal bylaws regulated the width of cart wheels to protect those expensive surfaces from wear.
The delivery route itself was, from very early on, also a branding exercise, which is a detail many histories of logistics omit. Toronto’s T. Eaton Company began its delivery service in 1870 with a single pony and a small wagon making two circuits of the city on weekdays and three on Saturdays. By the late nineteenth century, the firm had built a multi-storey stable behind its downtown store, and its red, white, and blue delivery wagons pulled by bay horses had become a recognizable signal of the Eaton’s brand, distinct from its rival, Simpsons, which ran grey teams. The consistency and visibility of a retailer’s delivery fleet was already, well over a century before the language of customer experience existed, understood as an extension of the retail brand itself.
Horse-drawn delivery also generated a body of municipal regulation and litigation that anticipated almost every concern that would later attach to motorized delivery: traffic accidents, public safety, animal welfare, noise, and the environmental burden of waste. Toronto bylaws, as early as 1834, prohibited horses from running or galloping immoderately on city streets, and courts spent decades adjudicating disputes over runaway horses, frightened animals, and liability for the resulting damage. The infrastructure of final delivery has always carried social costs alongside its economic benefits, and the horse era established that pattern long before the truck did.
Rural Free Delivery and the Reach of Government
While urban merchants were building horse-drawn route networks, a parallel and quite different last leg problem was unfolding in the North American countryside: the problem was not how to deliver efficiently, but whether delivery would happen at all. By 1890, nearly two-thirds of the American population still lived in rural areas, and while city dwellers had received free home mail delivery since 1863, rural households were required to travel to the post office themselves, often at real cost in time and effort.
Postmaster General John Wanamaker argued from 1891 onward that it made more sense, in both economic and social terms, to send a single mail carrier out along a route than to require dozens of rural families to travel into town individually to collect their own mail. Congress was slow to act; an 1892 bill to fund rural delivery died over its projected cost, and successive postmasters general regarded the experiment as financially impractical. It was not until October 1, 1896, that Rural Free Delivery actually began, on a modest experimental budget, in three towns in West Virginia. The experiment expanded rapidly. Within a year, 44 routes were operating across 29 states, and following a successful county-wide pilot in Carroll County, Maryland, in 1899, Rural Free Delivery became a permanent federal service on July 1, 1902.
The qualitative effect on rural households is well documented in the correspondence rural citizens sent to the Post Office Department at the time. One Arizona resident, writing while ill at home, expressed gratitude that mail now arrived at the door rather than requiring a three-and-a-half-mile journey into town, describing the change in terms of recognition and inclusion by the federal government rather than mere convenience. Rural Free Delivery is, in this sense, an early and instructive example of government stepping directly into a last-mile problem that private commercial delivery networks had no incentive to solve on their own, because the population density required to make a private route profitable simply did not exist in most rural areas.
The subsequent introduction of Parcel Post service in 1913 extended this logic considerably further. For the first time, packages rather than only letters could be mailed to rural addresses at affordable rates, and catalogue retailers such as Sears Roebuck and Montgomery Ward, who had spent decades cultivating a rural customer base they had no efficient way of physically supplying, suddenly had a delivery mechanism that matched their reach. Catalogue sales surged in the years that followed, and Rural Free Delivery, together with Parcel Post, stands as one of the more consequential government interventions in the history of final delivery in North America.
Canada followed a similar trajectory, on a slightly longer timeline. Free rural mail delivery began there on October 10, 1908, twelve years after the American experiment, following the same underlying logic: a federal carrier covering a fixed route was more efficient than thousands of individual households travelling to collect their own mail. Before that point, Canada had compensated for the absence of delivery with sheer density of fixed collection points rather than routes. By the turn of the century there was roughly one post office for every 5,000 Canadians, and on the Prairies these were frequently sited in grain elevator sheds spaced at roughly ten-mile intervals, the distance a fully loaded horse-and-grain wagon could comfortably cover in a single day’s round trip. The constraint was identical to the one driving Rural Free Delivery in the United States; only the solution differed, opting for proximity over mobility until the economics of free delivery made the older network of small post offices largely redundant.
The transition in Canada produced the same disruption to existing infrastructure that it had in the United States. The introduction of free rural delivery caused many of these small post offices, until then a genuine social hub for their communities, to close within a few years, as the same rail-based mail-car system already moving freight to depots was extended to sort and distribute mail directly along delivery routes rather than to fixed local offices. The most thorough academic account of this transition is Country Post: Rural Postal Service in Canada, 1880 to 1945 (University of Ottawa Press, 2003), an edited collection drawing on the recollections of twenty-eight Canadian postmasters, and the natural Canadian counterpart to Wayne Fuller’s standard American history of Rural Free Delivery.
The Truck and the Restructuring of the Final Leg
The internal combustion engine transformed final delivery with a speed that surprised many contemporaries. Motor trucks began displacing horse-drawn delivery on urban routes in the early 1920s, and by the 1930s, the transition was largely complete in most North American cities, although rural areas continued to rely on horses and rail feeders considerably longer. The number of registered trucks in the United States grew from a small base in 1910 to well over a million by 1920, a rate of adoption that reflected a genuine and immediate logistical advantage rather than mere novelty.
The truck did more than simply move faster than a horse-drawn wagon. It restructured the economics of the delivery route itself. A truck did not require feeding, stabling, or veterinary care, and it could carry substantially heavier loads over longer distances without the physical limits imposed by an animal’s stamina. Where horse-drawn routes had to be designed around the endurance and recovery needs of the animal, truck routes could be designed primarily around road networks, fuel availability, and the geography of the customer base. This was a genuine restructuring of last leg logistics, not merely an acceleration of the existing model, because it allowed a single driver to cover a substantially larger and more dispersed territory than had previously been practical.
The motorization of delivery also intersected directly with the legacy of the railway feeder problem described above. Where horse-drawn carts had once linked rail depots to their surrounding territory, that same function was now performed by trucks, but at a scale and radius the horse could never have matched. This expanded radius is part of what eventually allowed distribution to consolidate into fewer, larger depots serving wider territories, a structural shift that would prove important to the next major transformation in final delivery.
Motorization also supercharged the regular delivery route model that horse-drawn dairies, bakeries, and other merchants had pioneered. Through the 1920s, 1930s, and into the postwar decades, North American households received an extraordinary range of goods on regular truck-based routes, including milk, bread, ice, coal, and laundry. Many route drivers operated as independent contractors who owned their routes as transferable business assets, and the regular delivery route, organized around a known and stable customer base, represented one of the most economically efficient final-leg models North America had yet developed.
The Supermarket and the Reversal of the Final Leg
The most consequential disruption to that model came not from a new delivery technology, but from a new retail format that reversed the basic logic of final delivery altogether. The supermarket, which rose to dominance in North American retailing from the 1930s onward, was built on a principle that was the inverse of route delivery: rather than goods traveling to the consumer’s door, the consumer would travel to a large, centralized store offering substantially greater product variety at lower prices, achieved through volume purchasing and centralized distribution to the store itself rather than to the individual household.
This reversal depended on a separate and parallel infrastructure development: the spread of private automobile ownership. As car ownership expanded through the postwar decades, the time and inconvenience cost to the consumer of traveling to a centralized store fell considerably, while the cost advantage the supermarket could offer through volume buying grew. The weekly shopping trip gradually displaced the daily or weekly delivery route as the dominant provisioning model for most North American households, a transition that took decades to complete fully.
Home milk delivery is a particularly well-documented illustration of how slowly, and then how completely, this reversal occurred. In the early postwar period, the majority of North American households still received milk via home delivery. By the 1970s, that share had fallen dramatically, as supermarket milk, delivered to stores in bulk via the same truck-based logistics that had once served individual households directly, undercut home delivery decisively on price. The home delivery route did not disappear because it was technologically obsolete; it disappeared because the supermarket model, once automobile ownership made it broadly accessible, was simply a more efficient final-leg solution for a mass-market, price-sensitive consumer.
The supermarket era also drove a parallel transformation in the business-to-business final leg, the delivery of goods from distribution centres into the supermarkets themselves rather than to individual households. The standardized wooden pallet, developed and coordinated across industry in the postwar period with considerable influence from military logistics practice, together with the forklift and the pallet jack, rationalized the movement of large volumes of goods through warehouses and onto loading docks at a speed and scale that horse-drawn or even early truck-based handling could not approach. The loading dock itself became critical infrastructure, serving as the interface between the long-haul transport network and a building’s internal logistics. This solved the warehouse-to-dock portion of the supermarket’s final leg extremely efficiently, even as the dock-to-shelf portion and the final household-level leg from store to home remained comparatively labor-intensive.
Parcel Carriers and the Professionalization of Delivery
Alongside the rise of the supermarket, a separate and increasingly specialized delivery industry was professionalizing the parcel side of final-leg logistics. United Parcel Service, which began as a messenger company in the early twentieth century, developed over the following decades into a package delivery network distinguished by tightly optimized routing, standardized vehicles, and rigorous driver training. Federal Express, founded in the early 1970s on the premise that overnight delivery of high-value packages represented an underserved market, introduced the hub-and-spoke air logistics model that would later be widely adopted across the parcel delivery industry, effectively making geography a far less significant constraint on delivery time than it had been previously.
The labor dimension of this professionalized parcel industry is worth noting in its own right. Route drivers and parcel delivery workers became among the more organized segments of the broader logistics workforce, and unions such as the Teamsters negotiated contractual terms covering pay, route ownership, and working conditions that shaped the economics of delivery for decades. The professionalized, route-driven, and unionized parcel delivery model that emerged in the mid-twentieth century would remain the dominant structure of final-leg parcel logistics until the rise of electronic commerce began to reshape demand in the final years of the century.
E-Commerce and the Return of the Household as Destination
The commercial internet reversed the supermarket’s logic a second time, sending the final leg back toward the individual household after several decades in which retail consolidation had moved it toward the centralized store. From a logistics perspective, a supermarket is an extremely efficient delivery endpoint: a single bulk delivery on pallets can serve hundreds of customers, with the store’s own staff performing the final unpacking and shelving, and the consumer absorbing the genuinely final step of selecting a product and carrying it home. Electronic commerce disaggregates that efficient endpoint into a vast number of individual households, each requiring its own discrete delivery and each presenting its own access challenges, from locked buildings to absent recipients to inconsistent addressing.
Amazon’s introduction of two-day shipping under its Prime program in 2005 reset consumer expectations for delivery speed in a way that had cascading effects across the entire industry, converting what had briefly been a competitive advantage into a baseline expectation that required carriers to add routes, vehicles, and drivers at a pace that placed considerable strain on every part of the final-leg system.
Platform-based gig delivery services launched in the years that followed created an entirely new labour structure for urban final-leg logistics, one with notable structural echoes of the independent route drivers of the early twentieth century, but with one critical difference. The early route driver typically owned or leased a defined route with a known and recurring set of customers, a relationship that produced continuity and accountability. The platform-based gig worker, by contrast, is dispatched algorithmically to whichever delivery the system identifies at a given moment, with no lasting relationship to any particular route or customer. The efficiency gains of this model are real, but so is the loss of the route-level continuity that characterized final-leg delivery for the preceding century.
Electrification and the Search for the Next Model
The most recent chapter in this history concerns the electrification of delivery fleets and early experiments with autonomous and aerial delivery, developments still very much in progress rather than settled. Electric delivery vans and cargo bikes have become increasingly standard in urban final-leg operations, driven by a combination of regulatory pressure, falling battery costs, and the simple fact that the stop-start driving pattern typical of urban delivery routes suits current battery technology reasonably well. Drone and ground-robot delivery have moved from experimental pilots to limited commercial operation in specific, favourable environments, such as university campuses and selected suburban markets, though the technical and regulatory barriers to operating at full urban scale remain substantial.
What stands out across this entire history is not a single linear progression from inefficient to efficient final-leg delivery, but a recurring pattern of disruption and re-solution. The horse-drawn route gave way to the truck. The truck-based home delivery route gave way to the automobile-enabled supermarket. The supermarket’s centralized efficiency has, in part, given way again to the disaggregated household delivery of electronic commerce. Each transition was driven by a similar set of underlying forces: the relative cost of labor, the relative cost and availability of the energy or motive power involved, and the evolving expectations of the people the system ultimately serves. Those same three forces are visibly at work in the current period, and there is little reason, based on this history, to expect the present arrangement of final-leg logistics to be the last one.
Further Reading
Joanna Dean and Lucas Wilson, “Horse Power in the Modern City,” in R.W. Sandwell, ed., Powering Up Canada: A History of Power, Fuel, and Energy from 1600 (McGill-Queen’s University Press, 2016).
Clay McShane and Joel A. Tarr, The Horse in the City: Living Machines in the Nineteenth Century (Johns Hopkins University Press, 2007).
Ann Greene, Horses at Work: Harnessing Power in Industrial America (Harvard University Press, 2008).
USPS Historian’s Office, “Rural Free Delivery,” about.usps.com.
Wayne E. Fuller, RFD: The Changing Face of Rural America (Indiana University Press, 1964).
Amyot, Chantal, and John Willis. Country Post: Rural Postal Service in Canada, 1880 to 1945. Mercury Series. Ottawa: University of Ottawa Press, 2003.
Shane Hamilton, Supermarket USA: Food and Power in the Cold War Farms Race (Yale University Press, 2018).
“The Rise of the Supermarket,” American Heritage (1985).
Rose George, Ninety Percent of Everything: Inside Shipping, the Invisible Industry That Puts Clothes on Your Back, Gas in Your Car, and Food on Your Plate (Metropolitan Books, 2013).
Rick LeBlanc, EMBA, is the founder and editor of Reusable Packaging News (subscribe to the free newsletter) and editor of Western Pallet Magazine. A supply chain journalist with more than 30 years of experience, he specializes in pallets, reusable packaging, material handling and related operations. Rick is co-author of Pallets: A North American Perspective and Pallets & Progress: A Collected History of Pallets and Palletized Handling 1922–1945. He is also an advisory board member of the Virginia Tech Center for Packaging and Unit Load Design. Read Rick’s full bio.
