A world Without Email: Reimagining Work in an Age of Communication Overload

The Process Principle
The Power of Process
Early in my work on this book, I found myself browsing a little-used
shelf, buried deep in the stacks of Georgetown’s Lauinger Library,
filled with titles dissecting the nuances of industrial engineering. I
came across a collection of articles from a now defunct, early
twentieth-century business magazine called System, a publication
dedicated to case studies about the then new “scientific” approach to
management. These articles were almost universally breathless with
excitement about how much more money could be made by
industrial organizations once they began to think more
systematically about how their work was actually conducted. The
articles were also, it soon became clear, largely quite boring to the
modern reader. Scientific management in this era seemed to have a
lot to do with filling out forms in triplicate. System magazine loves
forms. In its pages, you’ll find pictures of forms, and you’ll learn
about their colors, how they’re perforated, even the material of the
folders that hold them (manila is preferred).
1
Hidden among these minutiae, however, I found a case study
from a 1916 issue that caught my attention. The subject was so old-
fashioned as to border on caricature: increasing the efficiency of the
brass works operating within the Pullman train car company’s
massive factory complex on Lake Calumet, fourteen miles south of
Chicago. But there was something about the story, written under the
direction of Pullman president John Runnells, that came across as
surprisingly contemporary. Many of Pullman’s thirty-three
departments depended on the brass works for key components,
keeping the roughly 350 men working the foundries and machine

---

tools in the brass works building perpetually busy. The problem with
their system for dealing with all this work, as the article explains, was
that there wasn’t really any system at all, but instead just a jumbled
mess of “slipshod methods.”
2
The brass department had only seven managers to help make
sense of the constant influx of work requests. These managers were,
of course, overwhelmed. As a result, everyone had to get informally
involved in managing the workflow. “In many places throughout the
plant, one man or another was devoting a part of his time to assist
the active seven,” it notes. “All the planning was being done
somewhere. And every man contributing by that much demoralized
his own particular work by the interruption.” As the article
elaborates, it became common for workers from other parts of the
factory to show up at the brass department and wait around most of
the day, bothering employees they knew, until they got the part they
needed.
In the first decades of the twentieth century, in other words, the
Pullman brass works had devolved into something that looks a lot
like the hyperactive hive mind workflow. However, unlike the many
knowledge work organizations of today that are suffering from a
similarly informal workflow, the leaders of Pullman, swept up in the
excitement of scientific management, were willing to experiment
with radical solutions.
—
To make the brass department more efficient, Pullman’s executive
team did something counterintuitive: they made its operation more
complicated. If you needed some brass work done, you now had to
submit an official form that contained all the relevant information.
To prevent employees from circumventing this process and reverting
to the more convenient status quo of informally bothering workers,
they literally locked the door and screened the windows. You now
had no choice but to use the newly enforced “regular channel.”
Once a request was submitted through a slot dedicated to this
purpose, it was subjected to a rigorous procedure. A clerk would be
tasked with figuring out a reasonable plan for accomplishing the
work, including what raw source materials would be needed and how
many worker hours were required to build the finished product. The
details of the plan were then distributed to the proper

---

subdepartments to ensure its timely execution. The specifics of the
process get intricate at this point, but also fascinating. Using armies
of clerks, the Pullman brass works seems to have replicated many of
the tasks that today we can implement instantaneously with a click of
a button on a computer application, implementing a sort of
steampunk IT system, made up of step-by-step instructions and
endless forms routed from desk to desk, like packets in a modern
network. They even built custom hardware, my favorite example
being an analog spreadsheet system that involved hanging brass tags
on a large wooden board divided into a grid, which somehow allowed
the work planners to quickly “cross-index” the current assignment of
workers to machines.
To implement this more structured workflow, Runnells had to
spend more money. There used to be seven administrative staff to
help organize the efforts of the 350 brass workers. Now there were
forty-seven. “Here is a vast increase in overhead,” the article admits.
Each of these new managers earned around $1,000 a year,
significantly increasing the department’s salary costs. “But does it
pay?” it asks. “It certainly has done so.” The new process dropped the
production cost of each train car by $100, which not only covered the
expense of the extra overhead but turned a “substantial profit.”
The article makes it clear why the additional overhead increased
profit. The old process—which was really no process at all—required
the 350 workers who actually produced the department’s valuable
output to constantly switch back and forth between informally
managing their workflow and actually executing the work itself. This
“demoralizing” double duty made them much slower at their actual
skilled efforts, reducing the return the department was getting from
its frontline workers.
When the workflow was restructured to largely eliminate this
double duty, the same workers could produce much more finished
brass in the same amount of time. “The old lack of method is and
never will be conducive to a betterment of standards,” the article
concludes. “But systemization promptly showed a surprising rise in
quality; workmen concentrated and the product showed the result.”
—
What industrial productivity hackers like John Runnells began to
discover in the first decades of the twentieth century is that efficiency

---

extends beyond the actions involved in physically manufacturing
something. Equally important is how you coordinate this work. The
problem with the Pullman brass works, in other words, was not that
the workers were bad at casting and polishing brass components, but
instead how these efforts were assigned and organized.
Like many fundamental ideas, it took a while for this new
mindset to take hold in the industrial sector. When Frederick
Winslow Taylor, the father of the scientific management revolution,
was first rising to prominence in the late 1890s, most of the energy in
this movement focused on the act of production itself. This is the era
that gave rise to the image of the draconian Taylorist consultant,
stopwatch in hand, trying to eliminate wasted motion on the factory
floor. Taylor himself made his reputation working with Bethlehem
Steel between 1898 and 1900, where, among other improvements, he
famously changed the style of shovel that workers used to move slag,
increasing the speed with which they could transfer the material
between piles. Pullman had integrated many of these ideas when
they built their factory during this period. John Runnells mentions
that the brass works had been laid out carefully with wide aisles, and
tools organized on racks, to increase the efficiency with which the
work was conducted. But as they learned, focusing only on physical
productivity wasn’t enough to get the department humming.
By the time the Pullman case study was published in 1916, a year
after Taylor’s death, magazines like System had increasingly
broadened their attention to include the information and decisions
that surrounded manual labor. The magazine was less about better
shovels and more about better forms to help figure out how much
shoveling needed to be done. To be more concrete, we’ll use the term