Dividing a task into its component operations lets the employer purchase exactly the amount of skill each step requires, rather than paying one skilled worker a high wage for the whole job, including its unskilled portions. The saving comes not from speed or dexterity but from matching the price of labour to the precise grade of skill each sub-task demands.
In pin manufacture, the price of each of the seven distinct operations is tallied separately, showing that the whole could be made far cheaper by assigning each step to a worker of the exact skill it needs. The accounting demonstrates that if one person performed all operations, the employer would overpay for the simple steps by applying the rate justified only by the hardest step.
Machines deliver their greatest economic advantage not through raw power but through their capacity to copy an operation with exactness and to repeat it indefinitely without fatigue or variation. Stamping, moulding, and printing illustrate that the cost of the original effort is spread across every identical copy produced thereafter.
The act of checking quality, counting output, and confirming that work was done honestly carries its own cost, which the organisation of a factory must minimise. Methods of self-registering tallies and arrangements that make fraud difficult are treated as economic devices, not merely moral ones.
A manufactory is examined as a quantifiable mechanism whose every operation can be measured, timed, and costed, so that improvement becomes a matter of observation and arithmetic rather than tradition or guesswork. Seeing production this way turns workshop practice into data that reveals where effort and expense are being wasted.
Whether it is economical to adopt a machine depends on the volume of articles over which its initial expense can be distributed; below a certain quantity, manual labour remains cheaper. The decision to mechanise is therefore a function of expected demand, not an absolute superiority of the machine.