A speculative incentive design: self-determined price commitments as a way of averting monopoly

post by mako yass (MakoYass) · 2020-04-28T07:44:52.440Z · LW · GW · 17 comments

Contents

  background
  the thing:
None
17 comments

[Epistemic status: A half-baked idea that I only came up with yesterday, but am nonetheless kind of excited by and I expect something like it will end up fitting in somewhere someday.]

background


Today, there is a lurking leech of a strategy where platform owners keep their margin high until the moment they need to fight off a competitor, then once a competitor emerges, they can easily lower it so that the competitor, with their scale disadvantage, can't win an audience, and is driven out of business. Tech is illegible enough that regulators couldn't reliably see through whatever lie is floated to explain the sudden price decrease, the evil can't truly be policed. The real damage done here is that the hero, the ascendant competitor, would only have been punished for trying, and I say "would", for this reason no hero actually steps up, most of this story doesn't play out, there is no competitor, so there is no price decrease.

There are many obvious examples of platforms we all rely on, which you and I know wouldn't cost anything at this point if competence, economy and generosity were well rewarded, whose maintainers are instead still taking monstrous cuts from all of the commerce that takes place on them. The gaming platform Steam, Uber Eats, Amazon Books, and so many that people outside of their industry don't even hear about, the hiring platforms, the licensed tools.

the thing:


These bad things couldn't happen if, before scaling up a technological platform with highly self-determined prices, platform owners are required to commit to all of their prices, many years in advance, and stick to them.

If a competitor arises who has reason to believe they can beat the committed prices, they can invest the money and enter the market, save in the confidence that the incumbent will not swoop down and take it from them once they're ready to launch.

For the first mover to avoid this, they must commit to prices that progressively decrease, in step with their well informed, insider analysis of the sectors expected propensity to create cheaper alternatives; and if they do it well, this obviates the need for anyone to actually do the redundant work of creating the cheaper alternatives! A company that produces the most convincing possible estimate of possible price decreases from fair competition over time will usually not face competition, nor should they, they will get only, yet all, the profits a fair and competitive market could support.

17 comments

Comments sorted by top scores.

comment by Dagon · 2020-04-28T19:00:30.649Z · LW(p) · GW(p)

This is a platform-killing limitation, if it were workable at all. There's no way to survive if you have to have prices you guessed at years ago, and newer competitors can pick theirs based on more recent data (including reaction to your committed prices).

Fortunately, it's not actually workable. Who is supposed to enforce this, and how will they determine which technological platforms are subject to this, and what changes in the platform justify different pricing? This implementation is more complex and difficult than the current SEC, FTC, and court system of identifying and limiting monopolistic practices.

I'm pretty convinced that interventions around pricing or post-monopoly behaviors are the wrong way to think about the problem. Look for ways to remove barriers to competition, not for ways to reduce the profit from non-competition (mostly; it's possible that a mix of approaches is better than any single one). To the extent that you have groups or agencies who can enforce anything in the first place, they should be enforcing data portability and transparency, in order for platform competition to flourish as customers change their preferences.

Oh, one more difficulty in price-setting enforcement: many MANY platforms don't have a single visible price. Ad sales are a very complicated contract + auction environment. Commissions have less-complicated but still variable pricing based on categories and contracts that change over time for key partners.

Replies from: MakoYass
comment by mako yass (MakoYass) · 2020-04-30T03:09:40.977Z · LW(p) · GW(p)

Though I don't especially disbelieve it, it would be helpful if you could tell some stories about how and why various platforms would been likely to be killed by (a decent implementation of!) pricing precommitment?

comment by Richard_Kennaway · 2020-04-28T09:38:22.284Z · LW(p) · GW(p)

How do you distinguish price-dropping to force the competition out of business (boo!) and price-dropping to be able to compete with the newcomer (yay!)? After all, cannot any competitive situation be interpreted as everyone trying to force everyone else out of business?

I remember that some years back, the price of several top-end commercial 3D modelling and animation tools suddenly collapsed. Houdini (edit: or maybe Maya) was around $6000 per year per user, and dropped to about $1200. Some trade fairs showcasing those high-end tools also stopped happening because the companies couldn't support the expense. The reason was that free tools like Blender, and cheap ones selling for a couple of hundred dollars, had over the years improved so far that a lot of professionals were deciding that they could no longer justify the cost of the expensive tools.

It was probably a good thing all round. The producers of the high-end tools have had to work harder to keep ahead of the low-end tools, and the producers of the low-end tools have reason to add more and more high-end functionality. And users at all levels have had better and better tools to work with.

I do not believe that any sort of regulatory interference with this would have benefitted anyone.

Replies from: MakoYass
comment by mako yass (MakoYass) · 2020-04-29T02:37:09.094Z · LW(p) · GW(p)

The problem is not the price decrease, the price decrease usually doesn't play out. Relatedly, self-determined price commitments aren't intended to prevent the price decrease that render competition non-viable, they would force it to happen.


The problem is the optionality of the price decrease. Situations where the incumbent can afford to reduce prices in order to disincent the creation of cheaper alternatives, the threat works, and so no cheaper alternatives arise, and so they don't.

So the proposal is to take away just the optionality of it. They can do whatever they want, but they must commit to it now, and being the market leaders, I think they often will have enough foresight that they can totally afford to do that. If they don't, maybe we'd all be better off if they weren't the market leaders!


I don't really see the story of Blender as a positive one. Afaik they only started receiving really adequate funding and industry adoption very recently, for the longest time being a fan of blender was kind of depressing, the industry mostly failed (and is probably still failing) to adequately reward its hero. self-determined price commitments would have resulted either in the emergence of a cheaper commercial competitor to maya or faster price decreases from them.

In the former case, Maya might have failed to anticipate competition and been driven bankrupt when it emerged. Whether that's terrible or not might depend mainly on how gracefully we can manage bankruptcy.. for physical assets they simply move into the hands of new owners, for digital assets I guess it must be more destructive, I imagine the fallout would hurt a lot of their customers?

Replies from: Richard_Kennaway
comment by Richard_Kennaway · 2020-04-29T08:05:34.766Z · LW(p) · GW(p)
I don't really see the story of Blender as a positive one. Afaik they only started receiving really adequate funding and industry adoption very recently, for the longest time being a fan of blender was kind of depressing, the industry mostly failed (and is probably still failing) to adequately reward its hero.

That looks like the same story I told, with a different framing. Free/cheap was snapping at the heels of the expensive software, and things evolved in a way that has benefitted everyone, all as a result of the users and companies taking their own decisions as they saw fit in pursuing their own interests. There is a wide spectrum of tools from free to expensive, and they all get better from year to year.

self-determined price commitments would have resulted either in the emergence of a cheaper commercial competitor to maya or faster price decreases from them.

I don't buy this, for all the reasons that Dagon has posted [LW(p) · GW(p)].

All the proposal does is reduce the ability of companies to respond to the conditions around them. It is indifferent to whether the responses they prohibit are ones you approve of or ones you do not.

comment by Ericf · 2020-04-28T13:06:04.101Z · LW(p) · GW(p)

Primary flaw is that you often can't buy the "same product" (in this case, a service) from one year to the next. So there is no way to enforce a future price, given that lirerally no-one knows exactly what the monopoly provider will be selling 2 years from now (just that it will fulfill the same need)

comment by ChristianKl · 2020-04-29T06:36:39.691Z · LW(p) · GW(p)

There are competitors to both Uber Eats and Amazon Books. It doesn't seem like this mechanism prevents people from trying.

On the other hand, if you force companies like Amazon that warehouse a lot of products and to commit to prices years in advance you destroy a lot of economic value. It's vital that an online seller can clear unbrought product by giving rebates to his customers.

Replies from: MakoYass
comment by mako yass (MakoYass) · 2020-04-30T01:46:52.883Z · LW(p) · GW(p)

Mm I guess they fail as examples (I am generally bad at coming up with examples of things though)

I think in that case the price commitment would apply to Amazon's (average?) markup, not individual product listing prices.

comment by cousin_it · 2020-04-28T08:54:17.262Z · LW(p) · GW(p)

I have another idea: if the mere existence of a competitor makes a monopoly drop prices all the way from monopoly price (way above break-even) to below break-even (necessary to crush the competitor), maybe the government should be selling some monopoly-prone goods at break-even. It would be very profitable for consumers and cost almost nothing.

Replies from: MakoYass
comment by mako yass (MakoYass) · 2020-04-28T09:40:40.888Z · LW(p) · GW(p)

They wont have to go below break-even to crush the competition due to the additional benefits of scale; there are things the incumbent will be able to afford to do that an upstart couldn't.


By what mechanism do you decide who runs the government-owned monopoly-prone productions, and why shouldn't it be a (bidding?) market with price commitments.

Whenever you propose a state-run alternative, you need to think about how they reliably hire competent people to run them. Can they ever really beat a mechanism that allows unknowns to step up and depose the incumbent the moment they can demonstrate that they can provide a more desirable product at a competitive price, because that seems like a pretty cool feature to me.

Replies from: cousin_it
comment by cousin_it · 2020-04-28T11:04:01.721Z · LW(p) · GW(p)

Now I feel a bit silly, because my comment wasn't a new idea at all, but rather the reason why public utilities exist. So maybe looking at their history and performance is the best way to answer your questions.

comment by mako yass (MakoYass) · 2020-04-28T08:18:45.879Z · LW(p) · GW(p)

An argument should probably made that reducing the profitability of tech even in the direction of fairness is bad, for anticapitalist realpolitik technocrat reasons. I'm not making that argument, but someone probably should.

Replies from: Pattern
comment by Pattern · 2020-04-28T19:29:36.508Z · LW(p) · GW(p)
An argument should probably made that reducing the profitability of tech even in the direction of fairness is bad

I mostly followed this part. (Fairness is vague.)

for anticapitalist realpolitik technocrat reasons.

But not this part.

Replies from: MakoYass
comment by mako yass (MakoYass) · 2020-04-29T03:58:27.776Z · LW(p) · GW(p)

Reasons within the genre of "capitalism is failing to adhere to capitalist virtues and that is good, because we are living in a very specific political reality where peoples' interests are best served by, channelling power towards illegible, fairly unaccountable people who make things"

comment by mako yass (MakoYass) · 2020-04-28T08:16:53.190Z · LW(p) · GW(p)

Hmm

Finally as at the EU level, a number of national competition authorities in Europe have been able to end investigations one excessive prices by receiving price commitments from the dominant firms investigated... [London Stock Exchange in 2004, Enel, Italian electricity incumbent in 2010, E.On regional gas suppliers in 2008

Is this the same kind of price commitment, I wonder? They sound like the kind of industries where there can't be competition even given all fairness, so probably not, probably the prices committed to were not self-determined. Not so exciting.

comment by mako yass (MakoYass) · 2020-04-28T07:55:42.504Z · LW(p) · GW(p)

I'm wondering if we can put more degrees of freedom in the conditions under which price commitment applies, meaning. I imagine this process being initiated by the market instead of being enforced on whatever arbitrarily defined category of product matches what we currently think it should apply to.

I feel like there is some dialogue between two honour-bound champions of industry that would lead to the incumbent being obligated to commit to a future price, simply as a result of answering questions and being held to their word. Probably involves some kind of wager somewhere.

Replies from: Pattern
comment by Pattern · 2020-04-28T19:32:52.544Z · LW(p) · GW(p)

Conditional prices might make more sense.

Tech is illegible enough that regulators couldn't reliably see through whatever lie is floated to explain the sudden price decrease, the evil can't truly be policed.

Though the "moral" aspect of this post was a bit much.