#797 1898 · Frederick Winslow Taylor / Bethlehem Steel · Industrial engineering / labor productivity

Bethlehem Steel's shovel gang used the same shovel for everything, so a scoop of ash weighed a few pounds and a scoop of ore weighed 30; Taylor didn't ask the men to dig faster, he changed the shovel.

问题

output looks like a discipline problem when the real constraint is an unmeasured physical mismatch between the tool and the material

背景

By the late 1890s, Bethlehem Steel's shoveling gang worked with whatever standard shovel was on hand regardless of what they were moving, so the same tool that scooped roughly 4 pounds of light rice coal also scooped 30 pounds of dense iron ore in a single load. Nobody had questioned whether the shovel itself, rather than the men's effort, was setting the pace of the work, and the standard management response to low output was simply to demand the workers move faster with the tools they already had.

Frederick Taylor, studying the gang's work with a stopwatch, suspected the real constraint wasn't human effort at all but an unmeasured physical variable: how much weight a shovel actually delivered per scoop, which swung wildly depending only on the material's density, not on any deliberate choice by management or worker.

换别人会怎么做

Push the shoveling gang to work harder or longer — the standard management response to low output, since it treats the shortfall as a motivation or discipline problem rather than questioning the tool itself. It leaves the shovel unchanged, which means workers moving dense ore are still overloaded on every scoop while workers moving light ash are still underloaded, so no amount of pressure closes the gap a mismatched tool created in the first place.

他们看到了什么

Taylor saw that output per worker wasn't actually measuring effort at all, it was measuring a physical accident: the same shovel delivered wildly different loads per scoop depending only on the material's density, so a worker on ore was straining under 30-pound loads while a worker on ash was barely working with 4-pound loads, neither of which was anywhere near what a human body could sustainably move all day. The fix wasn't extracting more effort, it was discovering the one number that actually mattered, the optimal load a first-class laborer could sustain across a full shift, roughly 21 pounds, and matching the tool to that constant for every material instead of using one shovel for everything.

那一手

Taylor determined that a first-class laborer could sustain his largest daily output shoveling loads of about 21 pounds regardless of the material, then had Bethlehem Steel stock roughly 8 to 10 differently sized shovels, a small scoop for dense ore, a large one for light ash, so that every material, whatever its density, was moved in loads close to that same optimal 21 pounds.

为什么管用

Human sustainable output per motion is a real physical constant, not an arbitrary target, so once Taylor measured it, he could work backward to the correct scoop size for any material's density and standardize on that instead of standardizing on a single shovel shape. Stocking 8 to 10 differently sized shovels cost Bethlehem Steel almost nothing next to the wages of the shoveling gang, yet it meant every worker, regardless of what they were moving, was operating at the same optimal load per scoop rather than being randomly over- or under-loaded by whatever material happened to be in front of them. That is why the fix produced gains no amount of exhortation could have: output per shoveler nearly quadrupled, the same tonnage needed roughly a third of the workforce, pay rose for the workers who remained, and cost per ton fell by more than half, all from changing the tool rather than the people.

值了多少

Over about three years, average daily output per shoveler rose from 16 tons to 59 tons, the workforce needed to move the same total tonnage shrank from between 400 and 600 workers to about 140, daily pay for the men who stayed rose from $1.15 to $1.88, and Bethlehem Steel's cost per ton shoveled fell from 7.2 cents to 3.3 cents, according to Taylor's own account in his 1911 The Principles of Scientific Management.

什么时候会失灵

The method depends on there being a genuine physical or procedural constant the existing tool or process is silently violating — if output is actually limited by something other than a measurable per-unit constraint, worker skill, equipment breakdowns, upstream supply, resizing the tool changes nothing. It also requires that the task be repetitive and physically homogeneous enough for a single optimal unit of effort to exist and be worth calibrating around; a job where the ideal load varies by more than material density, terrain, worker fatigue over a shift, individual physical variation, needs more than a set of fixed tool sizes to fully close the gap. And it depends on management actually being willing to invest in task-matched tools rather than continuing to treat low output as a discipline problem, since the entire insight is worthless if the diagnosis is made but the fix isn't funded.

后来呢

The shovel study became one of the founding case studies of scientific management, cited for over a century in industrial engineering as a clear demonstration that a fixed physical constant of human effort, not willpower or discipline, can be the real bottleneck in manual work, and that matching a tool's design to that constant can outperform any amount of pressure applied to a worker using the wrong-sized tool.

资料来源

[1]The Principles of Scientific ManagementFrederick Winslow Taylor, 1911 (Project Gutenberg, primary source), 1911gutenberg.org
[2]Frederick Winslow Taylor, the Patron Saint of the ShovelMental Floss, 2015mentalfloss.com