Randy asks us to elaborate on the empirical side of copying in the apparel industry, and rightly focuses on how symmetric, or asymmetric, copying is. As we describe in the paper, the fashion industry has many layers and roughly exhibits a pyramid shape. As one moves up the pyramid, prices and, broadly, design content go up. At the very top are the exquisitely-expensive custom clothes known as haute couture; at the bottom the mass market clothes one finds in Wal-Mart, Old Navy, and so forth. Writers on fashion have claimed that while this pyramid exists, fashion today (meaning in the last decade or two) has become far more "democratized," by which they mean that styles not only trickle down from the likes of Chanel and Prada but also bubble up from the streets and from cheaper labels. I tend to think this true.
That said, it remains the case that certain designers and labels are copied a lot. Some are simply more influential than others. So Randy is right to question whether there is symmetry in copying. But I think his analogy to pharma, while creative, doesnt hold up, in part because the claim we make, that "one is more likely, over time, to be a copyist than to be copied," is somewhat endogenous to the legal regime. In other words, since the legal regime is one of free appropriation, copying is easy, and firms copy those designs that seem to grab the marketplace. (Why one design gets hot and another doesnt remains a mystery for all concerned.)
In this scenario there are many copyists and one originator; over time, many many copyists and fewer originators. Imagine we had a similar free appropriation regime for pharma patents. We might also see everyone copying the blockbuster drug. But unless the orthodox account of IP is even more off-base than I thought, the drug wouldnt exist in the first place under this regime. And certainly future drugs would not come pouring forth. So I view the pharma case as more akin to a classic large-N PD, in which all but one of the actors would like to defect ex post and copy, but ex ante they agree to an IP regime that bars copying. Fashion is different, because the very act of copying the design kills it. "As fashion spreads, it goes to its doom," or something like that, was once declared by George Simmel.
In any event, Randy asks about timing as well. Runway shows do happen more or less simultaneously, and the fashion press nonetheless often identifies "themes" or trends across the shows. In part that might be the human capacity for pattern recognition (even where there is none) at work. But it also reflects larger cycles in design, "the zeitgeist," and so forth. (To see how fashion insiders describe their work, I can recommend the NYT, which has a lot of coverage, Women's Wear Daily, and Sascha Baron Cohen in the guise of "Bruno," on Da Ali G Show.) Most copying happens not across the same level of fashion (e.g. Gucci copies Marc Jacobs); instead those lower down the food chain tend to copy the more famous fish. There is even one firm, ABS, that essentially specializes in copying runway (and red carpet) dresses very quickly. And larger retailers like H&M and Zara have as their central business model the copying of design elements that appear on runway shows.
All of this is not to deny that much of the runway collection ends up for sale nowhere--the fashion shows are as much about generating press and buzz as actually showing designs that someone might one day buy. But the ideas surely get out there, and when some of them turn out to sell, the other firms move in to "reference" or pay homage to them.