Most economic textbooks take on a similar form. They start with an introduction that proposes a “production possibility” advantage to an economy. They make the assumption with two alternatives of production graphed with a convex curve and assume a better result than just specialising in either one. They justify this with an assertion that there are diminishing returns to labour and therefore avoiding specialising stops this.
They overlook the efficiency gains of large capital investments when specialising. Capital costs of manufacturing plant usually follow a power law which means the capital costs do not expand at the same rate as the plant capacity expands. In chemical plant the index of the power law varies. For aluminium production plant the power index is 0.80 meaning that to double the production capacity, the plant will cost 75% more. A caustic soda manufacturing plant will cost only 30% more for a plant of twice the capacity. Other chemical plants are in between these extremes. Doubling the plant size does not always mean doubling the labour force.
Do economists exaggerate the diminishing returns to labour? They depend on assumptions rather than collecting data. They never seem to know that in some cases larger material costs can receive a discounted cost. My thought is that they do not have experimental backing for their assertions. And this is only an introductory assertion. If economists are going to pretend to be scientific they had better do more measurements, stop assuming, and test their theories with real data.
When it comes to trade they produce the idea of “comparative advantage” which supports specialisation rather than the production possibility theory contradicting this phoney theory.