No End to the Economic Benefits of the Division of Labour

Chris Farris writing in The Exercise of Vital Powers (commentary on liberty, politics and economics) Blog (here) muses:

“Random thought as I was reading Greenspan’s book. Smith’s fundamental insight was that economic growth occurs with division of labor.

But how much can division of labor can the human mind support? Is there a limit to the complexity we can comprehend, beyond which division of labor ceases to have economic benefit? As we know more and more about less and less we eventually will know everything about nothing.”

Comment
Amusing thought, but misses the point. I am not into logic philosophy but I suppose there is a name for the logical error of extrapolating from a perceived (even measured) trend to arrive at a nonsense conclusion (and no doubt a reader will tell me what it is called).

The division of labour was certainly an idea (its origins preceded Adam Smith by centuries) of which he made great use of, and also pushed it further in the idea of it being limited by the ‘extent of the market’.

Unfortunately, most readers don’t bother reading past the famous pin-factory example (WN I.i.3: pp 14-15), which was widely known about at the time from Diderot’s ‘Encyclopedia’ and various other publications (see Jean-Louis Peaucelle: ‘Adam Smith’s use of multiple references for his pin-making example’, European Journal of the History of Economic Thought, 13:4, Dec 2006, pp 489-512).

If you read on a little further to page 22-23 of Wealth of Nations, you find Smith’s less-famous example of the effect of the international division of labour and specialization in the fabrication of the common woolen coat worn by day labourers. Taken together, you see why Chris Farris is wrong to worry about ‘knowing everything about nothing’.

The breaking down of some work tasks into their constituent parts reaches a physical limit but by separating out work tasks into specialised processes there is an enormous potential for further increases in productivity (increasing returns) which characterise the commercial mode of subsistence.

Even pin making is now made by one or two plants where in Smith’s days there were hundreds because automation replaced labour in pin-making. Now a few operators manage batches of machines doing all the 18-steps in the original process (but their machines are fabricated in scores of specialised plants located elsewhere, and their inputs are subject to increasing returns in their own right).

The division of labour is not confined artificially to one process in one factory. It is continually being propelled by innovation and adaptation across whole industries, quite separate from each other, that reduce unit costs in as many ways in products, that being incorporated in lower prices alter the processes wherever the products are inputs. Allyn Young (Economic Journal, 1928) described this process as ‘cumulative’ and it characterises modern manufacturing, and makes modern capitalism the productive force that it is.

Knowledge is dispersed, not reduced. The fact that there are four AER journals in place of one is not support for Chris Farris’s error; quite the opposite. Each author knows a great deal about a small specialist area, true, but together all authors know a great deal about more and more, not less and less. And this is only a single journal, admittedly one that specialises in mathematical abstractions about a fantasy state of general equilibrium, but there are thousands and thousands of other scholarly journals, which most people do not have time to read more than the titles of the articles in them. They don’t need to, but they know where to look if they did.

We are not in a state of ignorance; we would be if all the scientific work was absent.