Asymmetrical Information - Megan McArdle

10.16.12

What Would IP Free Industries Look Like? Probably Not Like Restaurants.

Yes, people can copy recipes without violating copyright. No, that doesn't mean restaurants are a good model for music or movies.

Over at the Volokh Conspiracy this week, the authors of the Knockoff Economy are blogging about their book, and their argument that imitation is the sincerest form of innovation.  A sample from the post on restaurant meals, in which they point out the way that successful innovations, like molten chocolate cake, tend to end up on menus throughout the country with no IP bar to copying:  

The bottom line is that in cuisine copying coexists comfortably with creativity. Now, as readers will surely notice, cuisine and fashion do not lend themselves fully to digitization, and so copying a dish is different than copying an Mp3 file. Of course, copying a recipe digitally is very similar to an Mp3, and in this comparison lies a few of the major points of the book. Let us just mention two. 

One, creative fields differ in many respects, ranging from cost of production to susceptibility to digital copying. Our IP system, by contrast, does not draw many distinctions, and to the degree it does, it distinguishes crudely. Two, copying can be blunted when a good is analog rather than digital. Recipes and songs can be digitized and copied very easily using current technology. But built food and its musical analogue, live performance, cannot. So one significant lesson from the world of cuisine (and high end bars, which we also discuss) is that creative industries can prosper even in the face of widespread copying by shifting revenue models toward hard to replicate goods, like restaurants (hard to copy) versus recipes (easier) in cuisine, or live rock shows (hard to copy) versus recordings (easy) in music. In these instances we’ve just provided, restaurants and rock concerts are both a sort of live performance, in which consumers look for and hopefully get an experience which is hard to copy. In the case of recipes and music recordings, these are more easily replicable products – especially recordings, which can be copied perfectly.  

I'm glad they bring this up, because you frequently hear industries with low IP appropriability (food, clothing) offered as models for industries that currently have relatively high IP appropriability (music, books, movies).  The problem is that the cases really aren't very parallel.  

Food has survived with low levels of appropriability because it's almost impossible to produce a meal from Lutece on a mass market scale.  Restaurant food, and especially high-end restaurant food, remains an incredibly labor intensive business.  For expensive restaurants, this is true throughout the supply chain:  they buy absolutely fresh, prime quality ingredients which have a very short shelf-life in their kitchen.  Applebee's may have its own molten chocolate cake, but it is not made with expensive, best quality chocolate and ultra-high fat cream whipped by hand; it probably comes parcooked and frozen from a central factory that has had to make an enormous number of compromises on taste and texture in order to de-skill the cooking and ensure shelf-stability and consistency.  So the inventor of a top-quality dish has relatively little fear of mass market competition.  

Similarly, couture clothiers have little reason to fear that target is going to knock off their multiple fittings and $500 a yard fabrics.  Except at the very high end (and sometimes even there) most of the cost of a garment is in labor, materials, shipping, and display, not design.  Knockoffs involve compromising on one or more of these items in order to offer a lower price, which means that the quality is virtually never so high. I myself do not think that anyone really needs a Birkin bag, but there is no such thing as a $40 identical copy of one.  

Even at the lower end, each designer makes choices about how garments will be fit and sized which means that you can almost never buy a straight knockoff of something you liked at another store; the knockoff might have been fit to a tall, curvy girl, while the original was constructed on someone more gamine.  

By contrast, in the industries where IP is becoming an issue--music, movies, and pharmaceuticals, and to a lesser extent, books--it is now possible to make endless perfect copies at a marginal cost of basically zero.  Without IP, classic economic tells us that the marginal price would naturally quickly fall towards zero.  There are in the real world numerous frictions which keep it somewhere north of nothing--but not necessarily all that far north, as far as I can tell.  

That is not just a difference in degree; it is a difference in kind.  (Which is why people indignantly asking whether it was okay for them to make mix tapes, or record off the radio, are not really proving as much as they think.)  I think the authors make an important point when they note that our IP system does not attempt to distinguish situations in which appropriability is naturally fairly high (like restaurant food) and ones where it is naturally very low, like prescription drugs.  Whether and how it should do so seems like a thorny question, but one important enough that we should be exploring it.

Of course, we could also phrase it the other way: maybe patent protection should be even stronger for especially low-appropriability products?  

Either way, it gives us a new way to think about increasingly strong calls for weakening or removing IP protections.  Moving towards an IP-free model implies radical things for the IP industries that we spend most of our time arguing about--for instance, in pharmaceuticals, unless we move to some sort of prize model, I think it implies the end of drug discovery.  I don't see how you finance FDA clinical trials through kickstarter.  

In music, it implies a return to a low-productivity artisinal profit model--a sort of deindustrial revolution.  Or rather, it implies that the low-productivity artisinal segment of the business will survive, while the rest of the industry is destroyed.  Live performance has, of course, always been with us.  

Has file sharing increased the number of venues?  Increased the number of concerts that are held in those venues?  Increased the price that people will pay per ticket?  Anecdotally, DC ticket prices for the bands I go to are about the same, adjusted for inflation--and occasionally, not adjusted for inflation--as they were in Philadelphia 20 years ago.  Concert tours aren't doing too badly, but is surprisingly dominated by a handful of new acts whose mean fan is about 12, and a bunch of 1980s and early 1990s acts whose mean fan is about 50. (And things that aren't concerts at all, like Blue Man Group and Cirque Du Soleil)  What was the prime concert-going demographic when I was in college now seems to be sitting them out.  

As far as I can tell, the much vaunted use of free music as advertising at best causes people to attend your concert instead of the one they would have gone to--it redistributes money between bands, but does not fix the supply.  You often hear a shift from recording revenue to concert revenue touted as a radical evolution towards a better business model, but to me it seems like a classic case of a price war that destroys the most profitable part of an industry, leaving low-profit discounters to scramble for a much-shrunken revenue stream.  The other suggested revenue streams, mostly licensing, are even worse, since of course, they themselves rely on IP protection of trademarks and t-shirt designs--or on other IP protected industries, like movies and television.  

Artisinal scale has other implications; an even younger industry, or perhaps a more amateur one.  The reason that you see so many old movies about broken down musicians and actors is that that used to be a real thing; performing six days a week and constantly travelling is not easy on the body, or for family formation.  Particularly in these days of child labor laws, when the authorities insist that your child spend the day in school instead of your dressing room.  

Indeed, you see this with high-end restauranteurs.  By forty, they're looking to get out of cooking: to be a franchise name with restaurants across the globe, or sell cookbooks, or have a television show.  That's because the sheer physical grind of standing on your feet in a hot kitchen for hours every night gets harder and harder to take as the body ages.  Growing up in New York, I met a fair number of people who had left restaurant kitchens to be private cooks, or something else entirely, but I did not meet a lot of old chefs who were still pushing out a hundred and fifty covers a night.  Early in their careers, chefs are mostly relying on their cooking skill, the ambience of their restaurants, and so forth to keep them solvent.  By the end of their career, more and more of them expect to rely on yes, strong IP protection to keep other people from making unlicensed copies of their work.

These are not necessarily arguments against looser IP, of course; perhaps the net gain outweighs any losses to musicians, or people who would have liked the music that poorer musicians don't create.  But a lot of the arguments for reducing IP protection make empirical claims about what would happen to the industry, based on parallels that don't seem very strong to me.  So it's good to see people tackling that fundamental difference head on.