Distribution of Agency

Note: This post is part of our League Symposium on inequality. You can read the introductory post for the Symposium here. To see a list of all posts in the Symposium so far, click here.

A Precis, provided by editorial staff (BlaiseP)

Inequality is not Unfairness. This (admittedly long) essay attempts to outline a trivial system in six cases and six – well, six and a half potential problems in this system. Statements and Axioms are presented, attempting to demonstrate how inequality arises from scaling and layers of abstraction, potentially creating instability.

*Solutions are discussed, culminating in the possible shooting down of my own argument, and a small speech on the subject of Aristotle’s Pursuit of Excellence.*

Introduction

“Fairness” is a tricky concept. It boils down to alignment with first principles, and it goes without saying that first principles are not universal. Thus, we’re unlikely to agree with a common definition of “fair”. I tried this before, discussing “justice”, but “fairness” is covered somewhat in that thread.

For this symposium, I won’t go down that rabbit hole. I’m specifically, intentionally, and with malice aforethought discounting fairness as part of my approach to this topic. For the beginning, anyway.

That’s right. No justice. No fairness. I am The Machine. I care not for your petty human desires and squabbles and inherited privileges. They are relevant only to the extent that sufficient “unfairness” leads you silly humans to aggregate antisocial behavior and violence.Then they matter quite a bit. Supposing that this level of unfairness is avoided, it is sufficient for my purposes. Spoiler alert: this comes up later.

What, if anything, is systemically wrong with inequality?

The study of economics (both macro- and micro-) is well outside the bounds of the page count for this post, and I must confess that it is not my particular area of expertise in any event. I have noticed, however, that the roots of economics aren’t really given much thought, even by economists.

This is not a dig on economists, as no profession is immune to this phenomenon. Nobel-prize-winning physicists usually are not scholars of the philosophy of science. Applied computational mathematicians are generally not experts in number theory, let alone metamathematics.

The point is that economists analyze economic systems in the context of the economic theory. They’re not thinking about the rules of the game, outside the context of the structure of the system itself… they’re thinking about the rules of the game, in the context of the structure itself… which comes with all sorts of assumptions that existed prior to the study of economics. Markets existed long before Adam Smith. You can’t study a system using the rules of the system itself, and economics is the study of the rules of the system, not the study of the system.

So let’s start by talking about the system. Permit me a very trivial and unfortunately long example, to illustrate six general categories of economic systems.

Two people, the Agents, live on a plot of land, the Playing Field. They have a small finite pile of seeds, a Scarce Resource. Each player takes two seeds out of the pile in turn. Turns take a quarter of an hour (big seeds!). To make things (computationally) easier on ourselves, we’ll assume that there are 20 hours in a day, and only 10 of them are available for actions of choice.

Now, these seeds are tasty, and chock full of nutrients and energy. They’re all you need, to live. It takes 1 seed to sustain an agent for a day if they are idle, and 3 seeds to sustain an agent for a day if they work. For every 3 seeds you plant, you will get a return of 10 seeds in 5 days. It takes one day of work to plant 30 seeds (3 seeds an hour), and one day of work to harvest 10 plants (1 plant per hour).

The tricky part is this: if I choose to plant, I’m away from the pile for an hour, so my counterpart gets four extra turns to take seeds out of the pile.

So, what happens?

Well, let’s say Agent Alice goes first, and Agent Bob goes second. Alice goes first, and takes two seeds out of the pile. This takes 15 minutes. Bob goes second, and takes two seeds out of the pile. We’re 30 minutes into our first day, nine and a half hours left.

Alice goes again, and takes another two seeds out (9 hrs, 15 minutes left). Now she’s getting hungry, so she eats one seed (needs 2 more at some point today) to get her energy going (zero time, this is already pretty complicated), and takes the remaining three seeds and goes to plant them. This takes her an hour, we will have left 8 hours and 15 minutes left when she gets back. Ah, but she’s gone right now, so Bob has fifteen minutes to take his turn, plus three more fifteeen minute periods before Alice gets back. Bob chooses to take 8 scarce resource out of the pile before Alice gets back.

Alice now returns. She has eaten a third of her daily quotient. Bob hasn’t had anything to eat (yet). Alice has planted a crop of 3 seeds, and Bob has planted nothing. Alice has no stockpile, and Bob has a stockpile of 8 scarce resources. What should Alice do next?

Well, herein lies the rub: without knowing how many seeds are in the pile of Scarce Resources, it’s impossible to even give a guess. If the pile is sufficiently small, Alice is already in trouble: she needs two more scarce resources just to survive the day. If the pile is gone, her choices are: die or cut a deal with Bob for two of his resources. Alice has nothing to trade except the seeds she just planted. Bob’s got her over a barrel and we’re back to the coercion thread.

That aside, what can we say from a systemic standpoint, regardless of the actual numbers we’ve given so far? Go back an replace all the hard numbers I gave with variables, have we learned anything yet? Why yes, we have.

First Case: Eden

In this case, the pile is big enough, or it is self-replenishing fast enough, that nobody needs to do anything except eat seeds out of the pile. Both Alice and Bob can survive with a trivial amount of time investment by any measure. Hurrah for Hedonism!

Second Case: Depletion

If the pile is sufficiently small, or the labor cost sufficiently high, or the time-to-yeild sufficiently long, we have another trivial result in the other direction: no matter what choices Bob or Alice make, the resources will all be gone (from consumption) before there are enough grown from a crop to keep even an idle person alive. This is an unsustainable economic engine that will collapse quickly: no matter what choices the agents make, they’re both doomed to starvation. There ain’t no justice and Nature is out to kill ya both. Boo! Malthusianism!

Third Case: Perfect Cooperation Required

Let’s look at the next case: that the number of initial scarce resources in the pile, the required labor-times, and the time-to-yeild are precisely sufficient for only a perfectly sustainable economic engine to take place. That is, let’s say that there exists exactly enough scarce resources available such that Alice and Bob can both plant enough scarce resources to enable them to survive and have enough left over to continue the crop. This is subsistence living at its most practically socialist: as long as they both cooperate fully, and they are both industrious, and nothing bad happens, they both live until their agency lifecycle reaches its durability limit. Given variability in the parameters I’ve given in this example, it’s even possible for this subsistence level to include idle time, or even oodles and oodles of idle time… depending upon how many seeds are in the original pile, how long it takes the “seeds” to mature, what the crop yeild is, and how much labor it takes to plant and harvest. You can see that it may be possible to get enough planted and have enough in stock such that you can have plenty of idle days between planting and harvesting to take up trivial or enjoyable non-working activities. Like, say, blogging. The important part to note is that if either agent diverges from the optimal strategy, they both starve to death. Depending upon how much idle time you get out of the deal, this may practically look like “16 Tons” or the life of The Feeders of Vaal, but the odds that you’ll wind up with some sort of authoritarianism are pretty high.

Fourth Case: Decoupled Cooperation Required

The fourth case is that the initial conditions are sufficient to support one set of suboptimal choices with sustainability. What do I mean by this? It means that there are sufficient resources in the pile, or crop times are short enough, or yeilds are high enough, that some cooperation is necessary for both Alice and Bob to survive, but this is independent of the other agent’s choices. We may need altruism. One of the two agents must spend less time idle than the other in order for the amount of resources to be sustainable, but (assuming one takes the burden) there are sufficient resources available to sustain both parties. This doesn’t even need to be a case of injustice, really, depending upon how this is worked out between the two agents… maybe Alice likes planting and harvesting and enjoys seeing Bob frolic shirtless in the daisies, so she takes on the planting and harvesting duties willingly. Who knows.

This incarnation, or one like it, is probably the case where most people argue about what is actually just. Most arguments about justice are disconnected from the actors.

Fifth Case: Full Independence, with rationality binding

In the fifth case, both agents have a sustainable existence, provided neither is self-destructive or malignant. Alice can plant and grow enough to keep herself alive, idle time notwithstanding. Bob can plant and grow enough to keep himself alive, provided Alice doesn’t burn down the fields and salt the earth. Here a threat of force can be binding.

Sixth Case: Full Independence, no rationality binding

In the last case, there are zero constraints on either party. No matter what Alice chooses to do, and no matter what Bob chooses to do – even if the other agent is malignant – both agents have a sustainable existence. The playing field is big enough, and/or the number of scare resources is large enough, that even should Bob go feral and try to hunt down Alice and murder her for kicks, she can survive and thrive independent of whatever Bob is doing. No threat of force is binding.

Summary of Cases

These six cases simplistically feature only two agents, and only one resource. It seems evident that one might have a different take on what “justice” is, in any given case above, depending on the case.  For example, the Libertarian members of the crowd might have a hard time with the third and fourth cases, but have pretty simple answers for the first, second, fourth, and fifth. I will not be surprised if the comment threads on the other posts in this symposium feature The League talking past each other on this point.

In the real complex world, there are 7 billion agents as actors.  Many things qualify as scarce resources, and some of those resources are more scarce than others.  Some of those resources are in fact tied to the scarcity of another resource: oil, clean water, and arable agricultural land are the big three.

What does this have to do with inequality?  Bear with me, we’re almost there. In a nutshell, resource scarcity is destabilizing.

I’ll now list the potential problems, exception scenarios, as ’twere. Doubtless we’ll see some sort of incarnation of one or more of these problems in the other posts in the symposium.  In fact, both Roger and Nob have already shown cases of these problems in the comment threads of their respective posts.

Problem Number One

Given the number of real-world variables involved, the problem is computationally open. By this I mean that there is no programmatically achievable result that can “solve” this set of equations for us. Not using any finite state computational device that can be built in this Universe, anyway. We can’t computationally exhaust Go, and this is much more complex than Go. There is no optimized strategy that can be derived; there is no “right” answer. Many people will refuse to believe that problem number one exists, but it does. What does this mean? It means The Machine (me) cannot offer a certain best-case scenario, only a selection of possibly least-bad ones.

Problem Number Two

Each of the cases we’ve listed has different initial conditions, and different outcomes based upon the first principles of the agents in question. We may (read: do) hold independent ideas of first principles that lead us to have divergent ideas of what ought to be done in each case, even if we all agree which case applies.

Problem Number Three

This is a particularly difficult problem.  We effectively use one proxy measurement (money) for all of the scarce resources, currency markets aside. This solves all sorts of problems associated with barter systems, which is why we use money in the first place. But the monetary value of each one of these scarce resources is driven not only by global supply and demand, but local variance in supply and demand. Local maxima and minima can be wildly different from the global average. Since those do not map directly onto each other, a shortage of one scarce resource may have a disproportionate effect on a locality in spite of the relative “fairness” of the global market.

When global oil futures are driven up by speculation on how much holiday travel Americans might regard as an achievable luxury good this summer, heating oil costs in southern Argentina go up. Suddenly some players are making economic decisions like all players are in the First Case, above… when some are effectively in the Second Case. This is a drawback of the global economy. Does it matter? Again, discounting justice, well… that largely depends on how bad this disjoin is between the players, and how much non-economic power the disadvantaged group can leverage on the advantaged one. And which group you’re in, of course.  But the answer is “of course it matters”, the question is “how much”.  If you think the local price of staple grains in northeastern Africa is of no moment to you sitting comfortably reading this blog post, you might want to reconsider.   I mean, seriously reconsider.

Problem Number Four

There are a hell of a lot more than one type scarce resource, and the utility function for “Scarcity” may wildly vary for all the individual agents. Dismal science!  For the average Joe who is unemployed in the United States, internet access is probably not a vital good.  If I’m unemployed, I can’t get a replacement job without it.  Such is the life of a systems administrator.

Problem Number Five

Externality can be shifted across case boundaries. If I want cheap iPods, and making iPods as cheaply as possible here in the U.S. means sticking the iPod factory on the Missouri River and dumping toxic sludge upstream of St. Louis, I have to deal with the people in St. Louis… or I can move the sludge-producing factory to Zimbabwe where food scarcity is a serious problem and people will take sludge in the water and a 35% greater chance of cancer over a lifetime… for seeds today.

As we’ve displayed on multiple threads in the past, who really knows better than the worker what conditions they’re willing to endure to make a buck?  Practically speaking, there’s not much we can do to influence other nations’ safety standards.  We don’t even do a really great job of our own.

Problem Number Six

Generally, these cases in practice are not closed. Once we add a sufficient number of additional agents, it may be the case that some set of agents (say, most/all Americans) fall in one case (we can assume we’re almost all effectively in one, five, or six, I would think)… while some other set of agents (say, some Africans) are in another case (the second, third, or fourth)… and some third set of agents (say, commodity traders) are interacting with both. Why is this a problem? Our intuitive respective idea of fairness – which is based upon our context as Americans – might not survive scrutiny when we acknowledge that the cases do not have closed boundaries. Provided everyone is eyes open, we can probably cope with this one here on the blog, but I would contend this is a serious problem, given our political system.

Problem Six Point Five

Also since these cases aren’t closed, they doesn’t consider the consequences of “Things Go Wrong”.

There are a goodly number of proxy rich, which is a final wrinkle worthy of an entire post in and of itself.

People who control, say, large mutual funds.  These exist for all sorts of reasons, but the unfortunate side effect is that the money put into these  “retirement vehicles” is decoupled from the people who rely on those retirement vehicles. Though the mutual fund managers issue bottom-line reports to the investors, multiple layers of abstraction tend to insulate the sources of capital from its effects.

There are people, right now, running organized campaigns against corporate entities that they themselves are (unknowingly) invested in supporting on the back end. Indeed, we are encouraged, arguably trained not pay attention to this money.

Let’s table problems number one, two, and four.  Let’s ignore half of six and the last wrinkle. I’m The Machine, remember. Just ignore them for the rest of this post.   Other people are undoubtedly covering those bases elsewhere, and even should we all agree that those potential problems  are not yet problems, we are left with serious concerns.

Statements

Liquid capital in the global economy will migrate across multiple cases. There are good reasons why this should be the case.  If we want Somalia’s economy to graduate from piracy, we need entrepreneurs to start businesses in Somalia

Consider these enormous incentives for ignoring moral hazard and externality:

• Most of our global economic agents (corporations) are proxies.
• The measure of utility (money) is also a proxy.
• Our own purchasing power is a function both of our population size and our wealth disparity.

A programmer would say, we’ve done a fine job of abstraction and encapsulation. The information is all hidden away in private members, featuring private and protected access methods for those who control the information and public methods for the public at large.

The downside is the actual value of “things” – for most of the players (as a number of independent agents) – is practically decoupled from their value – as represented in monetary terms – for most of the players in the game who control the vast majority of the wealth.

Liquid capital is going to follow profit incentives which are separated from our ideas of justice by layers of abstraction, independent of local variance from global supply and demand.  If we were truly rational, completely informed decision-makers, this would not matter.  Practically, if the consequences of all of that fall on other people, “The Machine is like Honey Badger, he don’t care.”

Axioms

Discontent Devolves to Anger

The less economic security a nation has, the more unstable it is.  The more unstable it is, the more likely it is to become a failed state.  Failed states engender less economic security in neighboring states.  Irritated members of failed states are known to take action against what they perceive as causes, independent of the actuality of causation.

Agent Abramovich can drop $785 million dollars on a yacht, and Agent Boris has trouble putting food on the table.  This might lead to troubles. If Abramovich really just wants a ridiculous wealth symbol. Boris might be discontent.  Yet, Boris might like to own that ridiculous wealth symbol himself someday, so there is a brake on Boris’ discontent.

Boris’ discontent turns into anger is when Abramovich is the controller of Gazprom and he hikes the price of fuel $0.50 a gallon, and he’s buying yachts.  Now Boris is having trouble filling up his delivery truck. Though a multitude of market forces beyond Abramovich’s control have led to the gas price hike, Boris will say, “I don’t mind you getting yours, but you’re killing me, doing it.”

Immobility

Moving your base labor requirements offshore makes your labor chain as complicated as your supply chain. It might be immediately cheaper, but the increase in complexity comes with an increase in brittleness. If your base labor requirements need to move, you have to do it all over again. This is hugely inefficient in terms of absolute skills maintenance.  If we were true technocrats, we would not bother providing machinist training to someone in some other country when we have a machinist that needs employment here.  Wouldn’t it make more sense to train that worker somewhere else to do some other job and leave the machinist jobs to the guys that have already spent 15 years becoming a master machinist?  Well, says the libertarian response to the technocrat, “because you want the market to encourage skill development where the wages are right”.  In a optimal world, sure… but the borders are not open, so the machinist here in the U.S. can’t just pack up and move to New Zealand to be a machinist there even supposing she wanted to do so, even supposing she could live like a duchess there.

Without open borders we are trapping workers with advanced skill sets in geopolitical containers that don’t have the jobs, while the global nature of the economy means jobs migrate to geopolitical containers that don’t have the skill.  Don’t misunderstand me, I’m an open border guy.  Unfortunately, I consider them to be less politically likely than Peace On Earth in my lifetime.

(For the record, I know that machinists are actually a pretty in-demand job here in the U.S.)

Here in the U.S., is it a serious problem, today? I’m not convinced that it is but I expect not.  Ward, for example, points to a feedback mechanism in that you can’t really outsource innovation. Just labor. That’s a valid point.

Focusing too much on an American-centric view in the here-and-now is missing two forests for the trees. From a global perspective, the ability of the vastly wealthy to be disproportionate drivers on the price index for scarce goods is a global issue, precisely because we have a global economy and the vast majority of the players are broke. Given the acceleration of this disparity, I call this a systemic problem with the global economy.

It will get worse in the immediate future.

When fifty bucks a month is living like a king somewhere, I can pay anybody there fifty bucks a month to make cars, and then ship the cars here. As long as the costs of transportation are low enough, there will be near zero demand for car manufacturing labor in the U.S.  If labor there is cheap enough and at-will employment is the norm, I can hire and fire my way through 20 employees abroad until I find someone who will be – in a year – as good as 70% of the automotive workforce here.  If there’s N x 20 people looking for a job there (for sufficient N), it will not take long to get a stable, reasonably competent workforce.  It might be an order of magnitude less efficient than the workforce here, but if they’re two orders of magnitude cheaper, I have cut costs, increased profit, and thus I get to smoke expensive cigars and drink good scotch and play at the country club and do all the other stereotypical things evil soulless corporate overlords can do.  The fact that an automotive innovation company needs innovative thinkers locally is independent of where the actual assembly gets done.

And without sufficient blue-collar jobs around in any geopolitical arena, it’s tough to have a sustainable local economy.  In a recession, agriculture will recover before manufacturing, and manufacturing will recover before the service sector.  If a sufficient number of your jobs are in the service sector, you get a very slow, painful, and sluggish recovery like the one we are undergoing right now.

True, labor in the U.S. to do other things. These jobs may require new skill sets, that are not readily achievable.  The cost of transition isn’t crippling, but it may be very high. It’s always been this way: one door closes and another opens in the work place.

Responses

I’ve said before, “Don’t confuse the fact that I think that there is a problem with the idea that I think there is a solution.”  A problem may not resist definition, but it may still resist a solution.

American political and economic agencies have responded in many different ways.

The supply-side response

Through subsidies and foreign aid programs, we’ve attempted to ensure a baseline of goods distribution.  The drawback to this approach is that this artificially deflates the value of the commodity in foreign localities.  It costs less to buy a frozen American chicken in Nigeria than it does to grow that chicken in Africa. Don’t expect Nigerians to start chicken farms. Between tariffs and subsidies, Haiti can’t make money exporting sugar to the USA.

The liquidity-side response

Just give away big bags of cash to other nations. Of course, the grateful initial recipients immediately export that cash to Switzerland. The local economy will never see a dime of that money. We don’t have any control over the agents of a sovereign nation. If they’re corrupt, well, what can we do?

A hybrid approach (“Give them food and pay them to farm anyway”) is theoretically possible, but is less intuitive than either of the previous two approaches and it likely is politically unfeasible.

The protectionism response

I’ll just assume for the sake of argument that nobody thinks this works any more, although it still occurs through political will here and elsewhere.

Nationalization (to some degree)

Oil countries do it (and it works so well)! We can make the barrier to entry for speculative commodity markets extremely high, or get rid of certain commodity markets altogether. There are a lot of wealthy people who will fight tooth and nail to keep the commodity markets as unregulated as they can, so this would be politically difficult. In addition, assuming it is successful, there will be inevitable attendant rent-seeking. A veritable welfare state for rent-seekers. Good plan.

None of those responses seems particularly awesome. Their respective side effects range from highly disagreeable to utterly intolerable. We use all of the above to one degree or another in an ad hoc oddball hybridization which is fairly messy. It doesn’t work, but it doesn’t not work, either.

Finally, you can tax high income earners, and high net value holders, to cut down on their huge liquidity disparity with the rest of “us schmucks”. They may be justly entitled to their money and capital; this doesn’t matter. It’s a problem of scale, not justice.

It seems improbable that money can have any sort of reasonably common value when 1,226 single entities can have a net worth larger than the GDP of more than a dozen nations (it gets worse on the poles).

Essentially, you’re doing this not for any reasons of justice, but just because our economic system, through its multiple layers of abstraction, highly encourage money – but not necessarily wealth – to percolate upwards.  The container is getting dangerously top-heavy.

Self-immolation

Now, for your entertainment, here’s where I shoot this whole essay in the foot.  The Machine is, after all, an idiot in direct proportion to his usefulness, which is to say, a very useful idiot.

Consider this perfectly valid, compelling counter-argument… which some of you may have started penning for the comment section even as you’ve been reading.

Though power may not be evenly distributed, that’s not necessarily a bad thing. If those with more power than others were guided by Aristotle’s “pursuit of excellence”, their vast powers could be applied to vastly excellent ends, freed up from the need for consensus-building and every man’s definition of fairness.

Taxing the rich folk for the purpose of reducing their relative power doesn’t put the money in a pile and burn it.  It just transfers it to yet another agency that operates under multiple layers of abstraction itself, the government. For all this wise talk about reducing our debt, politicians shall ever apply a revenue stream to the immediate interests of their own constituency.  Since one of the fundamental issues above is “there are too many layers of abstraction”, this itself seems like at best a dangerous potential “solution”.

We started “not quibbling about justice”, and now we are deep into the weeds of Justice and Excellence.

Less even than “excellence”, the group of very high capital controlling entities can be motivated by a large enough variety of desires to render this a cosmetic problem instead of a systemic one.  If I have a trillion dollars, but I spend it going to the Moon rather than trying to make a trillion and a half dollars in aggressive commodity market manipulation, society may benefit by that, overall. Technology advances from building arcologies on Mars will require advances in bioengineering and other technologies that will hugely benefit everyone on Spaceship Terra.  Few people remember Carnegie as the Steel Baron, they remember endowments of thousands of libraries, of the building of the Hooker Telescope on Mount Wilson and Carnegie Hall

Would anybody care about the rich, if the rich all were trying to get rid of worldwide malaria or colonizing the asteroid belt? I submit not so much.