Roger Sperberg talks about why computers and software (and anything that uses computers or software) gets cheaper over time. He then applies this thinking to alternate pricing strategies for the publishing industry.

I do owe you something and the question is is it more of the same on a faster cheaper platform, or an improved version with better features or simply emotionally satisfying in the current buying cycle. If the law of recapitulation holds, I have to give you the same thing and still sustain modifications in every subsequent version. Because the law of entropy holds for any communication system with a fixed energy budget, either the environment is becoming more sophisticated and can sustain the costs of modifications, or the environment collaspses and the market bifurcates. However, costs are not the only measure. There is a stickiness based on memory-driven patterns that create sustainable behaviors, even illogical ones.

I have the same songs on ever fewer disks. At some point, those disks become non-playable and I have to replace even more songs on fewer disks again. As a consumer, I accept that as the cost of maintaining my cultural desires. Replacement value is never zero.

Simulation of human intelligence based on AI has always been a tough problem because where based on logic, it fails to reckon that social bonding is based on emotional attachment, so the principle of rationality is a weak predictor of human and market behavior. We keep buying SUVs, and even if we stop for awhile as we did in the 70s, as soon as another Reagan emerges to tell us that unlimited consumption is our birthright, we’ll start buying them again just to fit in. In the short term, we will defy the energy budget and pray for it to rain oil. In the long term, we will create alternatives. Investing is a game of timing and opportunity.

The connundrum of market behavior is the willingness of the consumer to sacrifice necessity for popularity and convenience. As I said on my blog, my neighbor is out blowing the leaves off of his lawn. The kneepads help.

You’re right: It does cost something to make that nth copy. And package it. And distribute it or serve it from the web. And support the user. And models that assume the production costs go to zero have faulty assumptions, as you point out.

But I’m talking consumer mindset here, not producer economics. For shampoo, or gas, or paperback books, the price goes up every year (and sometimes despite that the quality goes down) and consumers accept that. They understand that it costs more to make those things and so the cost of the finished product has to go up.

Publishers have to be aware that their customers won’t be in that mindset when they buy electronic books, but in the mindset of consumers who’ve seen prices fall, continuously, in every regard, for 25 years, or since the onset of the personal computer. (Longer than that actually, but not really widespread enough to be noticed.) I remember thinking in 1978 what a great deal I got on my computer because for the same amount of money as my friend who bought in the spring, I got 64KB of memory and not just 16KB in the fall. That was more memory than the computers NASA used to send men to the moon. And just think what you could get today for $4,000.

And if you as a producer can’t come up with some way to give your customers additional features or additional product or something, your customers will abandon you. (Of course, I’m counting lower pricing as something “additional,” in this case additional value per dollar.)

We don’t expect poets to write ever better poems, or musicians ever better songs, but we do have rising expectations of software coders. Who today would accept a character-based application? It might do the job as well, or faster, but what user wouldn’t expect a GUI front-end?

No, code isn’t manufacturing, and markets aren’t homogeneous. But that’s your worry as a producer. From my perspective as a consumer, you owe me. If you’re electronic.

— Roger

For any shared code (say interfaces to State or Federal systems for administrative data), the business decision is who pays initial high costs and how are those amortized across markets of variant sizes. For main product code, this is relatively easy to do. For interfaces, it is harder given that they tend to be more custom and their market is more fragmented. Too often, your spreadsheet experts identify a market and make assumptions based on it having a smooth coherent solution surface. Later as you actually sell into that market, you discover it is really a lot of little mini-markets that overlap but are distinct; in other words, the closer you get to the ground, the bumpier it is. The overlaps are actually the hardest to price because where you may think they will converge on a fixed or limit attractor, they are actually strange and continue to orbit without recrossing.

So, the law in theory is good, but unreal in practice. You are treating code as if it were a manufacturing process, and that is an all too common mistake of naive modeling. Code management is more like farming or husbandry than manufacturing, at least, living code is. Otherwise, it is like selling packaged chickens.

At the publisher where I’m working, which provides legal, tax, insurance and accounting documents for research, after a certain amount of editorial work the documents go online. And from that point on, the company is less like a traditional publisher and more like a library.

I’m with you. I think traditional publishers ought to be more like a pay library and create a direct relationship with readers and offer all their books for a single fee. So what if they all read John Grisham — and increase his readership tenfold — then they’re going to go into those other titles. This is definitely a model I would go for.

I think it’s a good deal for both parties — more money for the publisher (and author), more books for the buck for the reader. Of course, the middlemen suffer. The book distributors aren’t needed anymore and go out of business, and bookstores become less like WalMart and more like antique stores.

And Darryl (at randomthoughts) –

I wouldn’t say I discovered this law, nor that I’m the first to write about it. Still I couldn’t find anything written about it, so I coined the label and the expressions of it in this article.

– Roger

This is a beautiful idea. I would sign up for that service in a heartbeat. I would even pay a monthly fee for access, and keep the subscription if I didn’t get around to reading a book that month.