Reading Time: 9 minutes
On a good day, a long-distance train in the late 19th century felt like a self-contained world. The conductor moved through crowded coaches like a ship’s captain, tickets in hand. In the gleaming dining car, stewards ushered travelers to tables laid with white linen and polished silver, selling not just meals but the idea of civilized travel at speed. Behind that smooth experience, however, railroads knew they were hemorrhaging money — not only to bandits along the line, but to the steady trickle of cash, food, and supplies disappearing inside their own cars.
That tension between hospitality and larceny is where Pinkerton’s National Detective Agency found its niche. Beginning in the 1850s, the Agency quietly became one of the railroads' most important tools for understanding what really happened once a train left the depot, long before anyone had coined the bloodless phrase "loss prevention."
Contracts and conductors
Pinkerton's railroad work began in 1855 with its first major contract. Allan Pinkerton already believed what the railroads were learning the hard way: their sprawling networks, tight schedules, and dispersed workforces created more opportunities for theft than any single company could monitor alone. Initially, the contract specified protection and investigations for theft.
After the Civil War, the Agency also started focusing on the people who handled money on nearly every passenger train: conductors. Special operatives were deployed to conduct formal “tests upon passenger conductors,” treating embezzlement on the rails as a solvable problem rather than an unavoidable cost of doing business. Turns out, it was.
Testing the “captains of the train”
The conductor was the railroad's on-board authority. He collected fares, checked tickets, enforced rules, and ensured every paying passenger was properly accounted for. He also worked alone, with no direct oversight. That combination of responsibility and autonomy made the conductor indispensable — and, from the railroad's perspective, a walking liability.
Pinkerton's answer was elegant: turn ordinary journeys into controlled experiments. Operatives rode trains in plain clothes, taking seats throughout the cars. Their job was to observe every fare collected and every ticket taken — where passengers boarded and left, whether cash was paid or a ticket presented, whether a receipt was issued, and how the conductor handled each transaction. Then they compared notes with the railroad's books.
If a conductor collected a cash fare from Chicago to some dusty station down the line, Pinkerton could see whether that fare appeared in the company's ledgers or had quietly vanished into a vest pocket. Over time, the Agency built detailed tables showing how much money certain conductors failed to report. In some early operations, the figures were startling: large percentages of cash collections — thousands of dollars — never reached the railroad's accounts, that is, until testing and discipline sharply reduced the losses.
Allan Pinkerton insisted these investigations be run by “honest minds” under a clear, fixed-fee arrangement. Operatives were salaried, not paid by the case, and were expected to give conductors the benefit of any genuine doubt. The aim was accuracy, not theater.
From station suppers to rolling restaurants
In rail’s early days, food was simple: the train stopped, everyone sprinted into a station dining room, quickly ate whatever was on offer, and tried not to choke when the conductor yelled: “All aboard.” It was fast, chaotic, and the food ranged from adequate to punishing.
As railroads matured, they realized they could differentiate themselves not just with speed and safety, but with comfort and cuisine. Beginning in the late 1860s, dedicated dining cars appeared on prominent routes. These were self-contained restaurants on wheels, with full kitchens, stewards, waiters, and elaborate menus. A good dining car became a selling point.
Railroads boasted of fine china, attentive service, and regional specialties served at 60 miles an hour.
For railroad executives, that luxury came with a problem. A dining car concentrated revenue in a very small space: meat, produce, and alcohol loaded at the commissary; meals and drinks sold for cash on board; leftovers and supplies off-loaded or discarded at the end of a run. At every stage, there were opportunities for waste, theft, and creative bookkeeping. Railroads quickly realized the same quiet testing Pinkerton had pioneered with conductors could be applied to this rolling restaurant business as well.
By the late 1920s, railroad dining cars operated at a significant loss on many major railroads. Yet they had to be maintained—cross-country travel demanded them, and the quality of dining service had become a major competitive differentiator between railroad companies.
Stewards, waiters, staff, and crew
The dining-car hierarchy began with the steward, who oversaw the entire operation: seating patrons, assigning waiters, issuing checks, and reconciling the day's business. Under him, waiters took orders and handled payments; chefs ran the kitchen; pantry staff looked after silver, linen, and dishware. Behind the scenes, a commissary received bulk supplies, portioned them, and sent stock out to individual cars.
In Pinkerton’s view, each point in that chain was both a control and a vulnerability. To understand how losses occurred, operatives had to see the system from end to end.
Patterns emerged immediately. There were low risk targets, the ones who stewards would attempt to defraud:
- Two women traveling together, especially young girls
- Two elderly women apparently unaccustomed to travel
- One woman traveling alone
- Elderly couples unfamiliar with dining car routines
Stewards avoided high risk targets: Single men traveling alone (usually), business travelers familiar with the service, and anyone who appeared detail-oriented.
Riding along, watching the numbers
For an operative assigned to a dining-car test, the work began like any other business trip: buy a ticket, board the train, and take a seat. He was instructed to behave like an ordinary traveler, to avoid making acquaintances, and never to spend more time around stations than a genuine passenger would.
In the coach, operatives noted fares paid: who boarded where, what they handed the conductor, whether they received a ticket or cash receipt, and whether the conductor properly canceled and retained each piece of paper. Unlike store testing or other investigative work where operatives could discreetly take notes, dining car investigations required near-photographic memory. An operative had to:
- Remember check numbers (often 6-8 digits long)
- Recall exact orders for multiple patrons
- Note prices and totals without writing anything down
- Observe crew interactions while appearing as a typical passenger
- Document kitchen operations while working as crew
- Recall conversations verbatim for later reporting
As soon as they could, operatives working dining cars rendered reports on Form #76.
In the dining car, the test had its own choreography. The operative would enter, accept the steward's seating, and observe how promptly a menu and a check were given. The operative would order specific combinations and quietly watch how each item did, or did not, appear on the check. After the meal, the operative scrutinized the total, paid in cash, and noticed whether the change matched the bill.
These single journeys became data points. When paired with the railroad’s own records and with multiple operatives riding at different times and places, they revealed the patterns that mattered.
One such pattern revealed low-risk targets, the ones stewards would attempt to defraud:
- Two women traveling together, especially young girls
- Two elderly women apparently unaccustomed to travel
- One woman traveling alone
- Elderly couples unfamiliar with dining car routines
Stewards avoided high-risk targets: Single men traveling alone (usually), business travelers familiar with the service, and anyone who appeared detail-oriented.
How schemes worked in practice
The manuals and sample reports from the 1920s show that Pinkerton did not treat dining-car losses as abstractions. The Agency trained its operatives to understand the specific ways stewards, waiters, and commissary staff might turn small irregularities into steady side income.
One recurring method was the manipulation of checks. In theory, every order began with a steward issuing a check in sequence, noting the waiter's number and the number of guests, then sending a copy to the kitchen while the original remained on the table. In practice, that process created opportunities.
“Switching” was among the most common schemes. A patron orders a fixed-price combination dinner — roast beef, potatoes, a side, rolls, and coffee. The steward collects payment, files the check, and sends the guest on their way. Later, another patron sits down and orders the same combination verbally. Instead of issuing a new check, a dishonest steward quietly reuses the original. The second payment never enters the records.
Variations multiplied. A patron might add dessert verbally after the check was already written. If the steward didn't add the item, the extra cash stayed in the car but never reached the railroad. On busy runs, some stewards dropped old checks on tables where they saw the same combination of dishes already served, betting that passengers would glance at the total, not the details.
Short-changing and overcharging had their own logic. A steward might blur a handwritten total or mis-extend a line item, then hand back lighter change when a large bill was offered. If the customer spotted the error, apologize and correct it. If not, the “mistake” became income.
Pinkerton’s response was to design tests that matched these mechanics. In some operations, three or four operatives would board the same train and, at intervals, order the same combination meal verbally. Later, they compared check numbers. If two or more had been given the same check for separate meals, they had evidence that checks were being reused.
When discovered, one steward admitted guilt and refunded $2,000.00.
Coach sales and commissary risks
Railroads also sold food and drink in the coaches, where a single waiter might carry trays of coffee, sandwiches, and bottled soda through crowded cars. Here, Pinkerton recognized another source of leakage: cups.
On some lines, the price of a hot drink included both coffee and the paper cup. By quietly retaining unused cups and reusing them on later trips, a waiter could under-report sales while pocketing part of the revenue. Small scheme, but repeated dozens of times a day across many trains, it added up.
One Pinkerton suggestion to a client was striking in its simplicity: require the use of straws for certain beverages. If each drink required a disposable straw, not just a cup, staff could no longer hide and reuse cups without revealing discrepancies in straw usage. A basic control, born from close observation, made one of the easiest manipulations far less attractive.
Behind the scenes, the commissary is an even bigger stage. This is the warehouse where meat, produce, canned goods, crockery, and linens arrive and get divided among cars. It is also where much value could evaporate before it ever reaches a table.
Pinkerton embedded operatives there, too. They watched deliveries: how meat was checked, how perishable stock was handled, how many bottles were left with each crew, and whether empties were returned and properly accounted for.
Their reports describe spoiled fish buried under rubbish rather than being written off correctly, meat that never quite matched its supposed weight. They noted when someone quietly carried a few steaks or specialty items out “for a friend” on a dining car. They paid attention when a load of ice cream was waved into cold storage without being checked. Sometimes the most revealing moments came when a delivery arrived and was waved through without being counted, or when no one could explain why a particular item was “over-ordered” again and again.
To railroad clients, these were not just colorful stories. They were a ledger of small, repeated decisions that add up to serious money. During one investigation, it was discovered that 110 pounds of beef tenderloin were “undercharged” and 31 lobsters and 7 cans of crab flakes were diverted.
From conductor tests to modern investigations
Pinkerton operatives stepped onto trains quietly, notebooks tucked away, prepared to ride, watch, and reconcile story with reality — a practice that continued well into the 20th century even as the rail experience changed dramatically. Modern corporate investigations still ask the same questions: where does value change hands, who controls the records, and where do blind spots let small losses become chronic bleeding? The technology has changed—analytics platforms instead of paper checks, cameras instead of plain-clothes operatives. But the discipline of following the money, step by quiet step, remains the same.
