More Machines, Better Machines…or Better Workers?

How much of the rapid growth in output per man-hour in nineteenth-century cotton weaving arose from technical change and how much arose from price-driven substitution of capital for labor? Using an engineering production function, I find that factor price changes account for little of the growth in output per man-hour. However, much of the growth and most of the apparent labor-saving bias arose not from inventions, but from improved labor quality—better workers spent less time monitoring the looms. Labor quality played a critical role in the persistent association between economic growth and capital deepening in this important sector.