But Hayek is remarkably incurious about the actual social processes through which markets work, and in particular the forms of standardization that are necessary to make long-distance trade work. I imagine Scott’s counterblast going something like this: It is all very fine to say that markets provide a means to communicate tacit knowledge, and it is even true of many markets, especially small scale ones with participants who know each other, know the product and so on. But global markets do not rely on tacit knowledge. They rely on standardization – the homogenization of products so that they can be lumped under the appropriate heading within a set of standard codified categories. Far from communicating tacit knowledge, the price system (and the codified standards that underlie it) destroys it systematically.
Here is much more, mostly on James Scott.
I agree with the frustration about a lack of detail about markets. Nonetheless, I am more optimistic than this. Standardized products often fund a more general trade or communications infrastructure which the non-standardized products then piggyback upon. Think of funding a port with oil shipments (by the way, how many grades of crude are there?) but at the margin using it to trade strange toys as well. Alternatively, a highly standardized product can be communicating the (sometimes tacit) knowledge that liquidity and interchangeability and lots of direct competition matter more than does product diversity.
The Grossman-Stiglitz framework helps us think through the trade-off between the average and the margin. Let's say that some trades shift into the more standardized, liquid market and out of the more idiosyncratic market. Some knowledge is lost. But at the margin, there is now a stronger incentive for information-gathering, or knowledge mobilization, in the less liquid market. It will be easier to beat the rest of the market (price is not much of a sufficient statistic) and so people will try harder.
This topic is related to the current controversy over whether swaps contract will be overly customized (wider spreads and harder to monitor and higher regulatory risk?) or overly standardized (more liquid but less useful?).